universita' degli studi di padovapaduaresearch.cab.unipd.it/148/1/thesis_ortegac.pdfmembrane...
TRANSCRIPT
- 1 -
UNIVERSITA' DEGLI STUDI DI
PADOVA
Sede Amministrativa: Università degli Studi di Padova
Dipartimento di Chimica Biologica
SCUOLA DI DOTTORATO DI RICERCA IN BIOCHIMICA E
BIOTECNOLOGIE
INDIRIZZO: BIOCHIMICA E BIOFISICA
CICLO XX
TITOLO TESI
Functional properties of an unusual isoform of the
plasma membrane calcium ATPase: PMCA 2
Direttore della Scuola: Ch.mo Prof. Lorenzo Pinna
Supervisore: Ch.mo Prof. Ernesto Carafoli
Dottoranda: Claudia Karina Ortega
- 2 -
31 gennaio 2008
A papá
- 3 -
Hay bajo el sol un momento para todo,
y un tiempo para hacer cada cosa:
Tiempo para nacer, y tiempo para morir;
tiempo para plantar, y tiempo para arrancar lo plantado;
tiempo para matar y tiempo para curar;
tiempo para demoler y tiempo para edificar;
tiempo para llorar y tiempo para reír;
tiempo para gemir y tiempo para bailar;
tiempo para lanzar piedras y tiempo para recogerlas;
tiempo para los abrazos y tiempo para abstenerse de ellos;
tiempo para buscar y tiempo para perder;
tiempo para conservar y tiempo para tirar fuera;
tiempo para rasgar y tiempo para coser;
tiempo para callarse y tiempo para hablar;
tiempo para amar y tiempo para odiar;
tiempo para la guerra y tiempo para la paz
- 4 -
SUMMARY The plasma membrane calcium ATP-ase (PMCA) represents a primary system for the
specific extrusion of calcium from eukaryotic cells. Together with Na+/Ca2+ exchangers,
it is the major plasma membrane transport system responsible for the long-term
regulation of the resting intracellular calcium concentration. The PMCAs contain 10
membrane-spanning segments and the NH2 and COOH termini are both located on the
cytosolic side of the membrane.
The PMCA pump is the product of 4 separate genes, isoform diversity being further
increased by a complex pattern of alternative splicing of the primary transcripts. The
four basic gene products (PMCA 1-4) and the numerous splice variants vary in
expression level during development, have peculiar distribution in tissues and within
cells, and differ with respect to functional parameters, especially those involving
regulation properties. PMCA 1 and 4 are ubiquitously expressed while 2 and 3 are
mostly found in the central nervous system and, in lesser amounts, in the striated
muscle.
Alternative splicing is peculiarly complex in PMCA2 because it involves the insertion
of up to three novel exons at site A (variant w) and of two at site C (variant a). The
insert at site C creates instead a novel stop codon, leading to the truncation of the pump.
The site-C insertions eliminate approximately half of the calmodulin binding domain;
those at site A occur next to a domain that binds activatory acidic phospholipids.
Calcium enters the stereocilia of hair cells through mechanoelectrical transduction
channels opened by the deflection of the hair bundle and is exported back to endolymph
by an unusual splicing isoform (wa) of plasma-membrane calcium-pump isoform 2
(PMCA2).
Ablation or missense mutations of the pump cause deafness, as described for the G283S
mutation in the deafwaddler (dfw) mouse. A deafness-inducing missense mutation of
PMCA2 (G293S) has been identified in a human family. The family also was screened
for mutations in cadherin 23, which accentuated hearing loss in a previously described
human family with a PMCA2 mutation. The wa variant and their mutations were
overexpressed in Sf9 cells. At variance with the other PMCA2 isoforms, it became
activated only marginally when exposed to a Ca2+ pulse. The defect is more pronounced
in the dfw mutant than in the G293S (human) mutant. A third mutation was analyzed:
- 5 -
Oblivion mouse mutation (where a Serine is replaced by a Phenylalanine at position 877
in TM6), it was slightly more active than the pumps bearing the 283 and 293 mutations
but still much lower that of the wild type wa.
The other important aspect studied, it is the regulation mechanism of the PMCA 2 wa
variant. Microsomal membranes isolated from CHO cells transfected with PMCA 2 zb,
PMCA 2 wb, PMCA 2 wa and PMCA 2 wa (Tommy), where a mutation in the ATP
binding site occurs in the valine 586, were assayed. Both PMCA 2 zb and PMCA 2 wb
had higher basal activity than PMCA 2 wa. PMCA 2 wa (Tommy) only had about 25
per cent of the activity of PMCA 2 wa. In the presence of Calmodulin, PMCA2 zb and
wb showed the highest response to it i.e., the activity of these isoforms was over eight
times higher than of PMCA 2 wa. PMCA2 wa had lower affinity for calmodulin than
PMCA2 zb and wb, but had still higher affinity than the corresponding Tommy mutant.
The mechanism by which different phospholipids activate the PMCAs is not known.
However, given the location of the lipid-binding sequences in the pump, one speculated
that the interaction of acidic phospholipids with the calmodulin binding domain leads to
some structural rearrangement that weakens the autoinhibitory intramolecular
interactions formed by the COOH-terminal tail. The PMCA 2 zb and wb, when
overexpressed in CHO cells, had the same response to this acidic phospholipids
implying that the inserts next to the N-terminal phospholipids binding domain (variant
w) was not disturbed by the splicing insert. The PMCA wa was absolutely not
stimulated by phosphatidyl serine apparently, then, no activation by acidic
phospholipids can occur when the C-terminal binding site is undisturbed. The response
of the PMCA 2 b pumps to PS was 4 fold higher than that of the wa variant.
To investigate whether Calmodulin indeed removes its binding domain from the
cytosolic loops of the pump, we have introduced a mutation in one of the two sites that
interact with the CaM binding domain in the cytosolic loops of the pump. Prior to this,
the PMCA pump was modeled to decide the residue that should have been mutated.
From the model and previous works we decided to mutate the methionine 265 in
Alanine (M265A). The activity of the PMCA 2 zb and M265A pumps expressed in
transfected CHO cells evaluated as a function of Ca2+ in both the presence and absence
of calmodulin shows that the M265A mutant has a response to calmodulin which was
the same as that of the wild type at saturating concentrations of calmodulin. The
activation by CaM was higher in the M265A pump with respect to wild type pump at
1nM, 5nM and 15 nM calmodulin concentrations. This suggests that the autoinhibitory
- 6 -
sequence was bound with less affinity to the intramolecular binding site in the Met265
mutant as less calmodulin was required to relieve the inhibition. Evidently, the
inhibition of Met 265 decreased the affinity of the C-terminal autoinhibitory domain,
making it easier for calmodulin to remove it from binding site.
- 7 -
RIASSUNTO
Le Ca2+-ATPasi della membrana plasmatica (PMCA) sono responsabili dell'espulsione
specifica del Ca2+ dal citoplasma nelle cellule eucariotiche. Assieme allo scambiatore
Na+/Ca2+ rappresentano il meccanismo di trasporto più importante per mantenere la
concentrazione del Ca2+ intracellulare a livelli basali. Le pompe PMCA contengono 10
domini transmembrana e presentano le estremità amminoterminale e carbossiterminale
rivolte verso il citosol.
Nelle cellule eucariotiche sono presenti 4 diverse isoforme di PMCA, denominate
PMCA1, 2, 3, 4 e codificate da geni distinti. Il numero delle isoforme è ulteriormente
aumentato da meccanismi di splicing alternativo del trascritto primario di ciascuno di
questi geni. Le quattro isoforme e le numerose varianti di splicing mostrano un livello di
espressione che varia durante lo sviluppo embrionale, una distribuzione tessuto-
specifica e presentano caratteristiche biochimiche e funzionali proprie. Le isoforme 1 e
4 sono ubiquitarie mentre le isoforme 2 e 3 sono prevalentemente espresse nel sistema
nervoso centrale e a livelli minori nel muscolo scheletrico.
Lo splicing alternativo e un processo abbastanza complesso nella PMCA 2 perché
richiede l’inserzione de tre nuovi esoni nel sito A (variante w) e di due nel sito C
(variante a). L’inserzione nel sito C genera un nuovo codone di stop, provocando la
troncatura della pompa. Le inserzioni al sito C elimina approssimativamente meta del
dominio che lega la calmodulina; in vece quelli nel sito A, sono prossimi al dominio che
lega i fosfolipidi acidi.
Il calcio entra nelle stereocilie delle cellule ciliate attraverso l’apertura dei canali di
meccanotrasduzione grazie alla deflessione della stereocilia ed è esportato alla endolinfa
attraverso di una variante di splicing particolare (wa) della Ca2+-ATPasi della
membrana plasmatica 2 (PMCA 2).
Ablazioni e mutazioni missense della pompa provocano sordità, come e descritto per la
mutazione G283S nel topo deafwaddler (dfw). Una mutazione missense della PMCA 2
(G293S) che induce sordità e stata identificata in una famiglia umana. La famiglia fu
esaminata anche per mutazioni nella caderina 23, la quale accentuò la perdita dell’udito
nella famiglia previamente descritta associata alla mutazione della PMCA 2. La variante
wa e sui mutanti furono sovraespresse in cellule Sf9. A differenza delle altre varianti
della PMCA, wa è attivata solo marginalmente quando e sposta a polsi di calcio.
L’effetto e ancora più evidente nel mutante dfw che nel mutante umano G293S. Una
- 8 -
terza mutazione fu analizzata: la mutazione del topo denominata Oblivion (dove una
serina e sostituita per una fenilalanina nella posizione 877 nel dominio di
transmembrana 6), questa fu leggermente più attiva delle pompe che portano le
mutazioni 283 e 293 pero ancora molto piu basso de quelle del wild type wa.
Un altro aspetto importante studiato, e il meccanismo de regolazione della variante
PMCA 2 wa. Membrane isolate de cellule CHO trasfettate con PMCA 2 zb, PMCA 2
wb, PMCA 2 wa e PMCA 2 wa (Tommy), dove apparve una mutazione nel sito de
legame del ATP nella valina 586 furono studiata. Tanto PMCA 2 zb e PMCA 2 wb
hanno un’attività basale più alta in confronto alla variante PMCA wa. PMCA 2 wa
Tommy ha soltanto il 25 per cento de attività in confronto alla variante wa. In presenza
di calmodulina, PMCA 2 zb e PMCA 2 wb mostrano una risposta più elevata, l’attività
de queste varianti fu di otto volto maggiore a quella della variante wa. PMCA 2 wa ha
una bassa affinità per la calmodulina probabilmente basato nel fatto che manca de meta
del dominio che la lega. L’ affinità per il calcio della variante wa e comunque più alta a
quella del mutante Tommy.
Il meccanismo mediante il quale diversi fosfolipidi attivano le PMCAs non e ancora
noto. Tuttavia, data la localizzazione delle sequenze che legano i lipidi nella pompa, si
può speculare che le interazioni dei fosfolipidi con il dominio che lega la calmodulina
indurre a cambiamenti strutturali che indeboliscono le interazioni intramolecolari con la
porzione C-terminale della pompa. PMCA 2 zb e PMCA 2 wb hanno la stessa risposta
ai fosfolipidi acidi significando che l’inserzione prossima al dominio che legano i
fosfolipidi (variante w) non furono disturbate per le inserzioni di splicing. The PMCA 2
wa non fu assolutamente stimolata per la fosfatidilserina (PS), per tanto, la non
attivazione può verificarsi forse, quando il sito di legame C-terminale e indisturbato. La
risposta delle varianti b alla PS fu quattro volte in confronto alla variante wa.
Per investigare sì la Calmodulina davvero rimuove il suo sito de legame dei loops
citoplasmatici della pompa, avviamo introdotto una mutazione in uno dei siti che
interagiscono con il dominio che lega calmodulina che attua inibendo la pompa. Prima
di questo, la PMCA fu modellata per decidere che residuo doveva essere mutato. Del
analizzi del modello e di lavori precedenti, fu deciso di mutare la metionina 265 in
alanina. L’attività della pompa wild type (PMCA 2 zb) e del mutante (M265A) fu
analizzata in presenza e assenza di calmodulina. A concentrazioni saturanti di
calmodulina tutti due le pompe hanno la stessa risposta al nominato attivatore. A bassa
concentrazione di calmodulina (1nM, 5nM e 15 nM), il mutante sembra essere più
- 9 -
attivo del wild type. Questo potrebbe suggerire che la sequenza che inibisce la pompa se
lega con una minore affinità nel mutante e una concentrazione minore di calmodulina e
necessaria per liberare la sequenza che lega la calmodulina ai siti di rispettivi di legame.
Evidentemente l’inibizione della M265 diminuisce l’affinità al dominio autoinibitorio
C-terminale.
- 10 -
TABLE OF CONTENTS
ABBREVIATIONS .................................................................................................. 14
INTRODUCTION.................................................................................................... 16
1- General principles of Ca2+ signaling...................................................................... 18
2- Ca2+ sensors ........................................................................................................... 19 3- Ca2+-transporting protein ....................................................................................... 21 4- The plasma membrane calcium pump(PMCA ) .................................................... 27 4.1- Isolation and purification of the Calcium Pumps .............................................. 28 4.2-Cloning of the pumps and recognition of the isoforms ....................................... 29 4.3- The Plasma Membrane Calcium Pumps: Structural and Regulatory Characteristics ......................................................... 32 4.4- Isoforms of the PMCA pump ............................................................................. 33
4.5- Splicing variants ................................................................................................. 36 4.6- Protein interactors of PMCA pumps.................................................................. 39 4.7. PMCA pump and diseases ................................................................................. 40 4.7.1- PMCA pump knock out mice .......................................................................... 40 4.7.2- Hereadity deafness and other diseases conditions .......................................... 42
EXPERIMENTAL PROCEDURES ...................................................................... 44
1- Baculovirus Expression System ............................................................................ 44
1.1- Baculovirus Expression vectors.......................................................................... 45 1.2- Cytopathogenesis……………………………………………………………… 45 1.3- Generating recombinat virus by Homologous Recombination........................... 46 1.4- Generatin recombinat virus by site-specific transpositoion................................ 47 2- Cloning and virus generation of PMCA zb and wa variants.................................. 48
2.1- Construction of the recombinant baculovirus..................................................... 49 2.2- Plasmid DNA isolation from bacteria. Miniprep................................................ 49 2.3- Plasmid DNA isolation from bacteria. Maxiprep………………………………50 2.4- DNA digestion and agarose gel electrphoresis ................................................... 50 2.5- DNA fragments purification from agarose gel ................................................... 51 2.6- DNA ligation ...................................................................................................... 51 2.7- Transformation of competent E.coli cells........................................................... 51 2.8- DNA sequencing................................................................................................. 51 2.9- Generation of Recombinant Bacmid................................................................... 52 2.10- Transformation of DH10Bac E.coli cells ......................................................... 52
2.11- Analysis of recombinant DNA by PCR............................................................ 54 2.12- Transfection of insect cells ............................................................................... 55 2.13- Virus amplification ........................................................................................... 56 3- Expression and purification of PMCA 2 zb and wa variants................................. 56 3.1- Infection and expression of PMCA 2 in insect cells.......................................... 56 3.2- Sf9 cell membrane preparation .......................................................................... 57 3.2- Solubilization of PMCA 2 proteins .................................................................... 58 3.4- Calmodulin Affinity Chromatography ............................................................... 58 3.5- His-Tag affinity chromatograph ......................................................................... 58 3.6- PMCA 2 preparation for SDS-PAGE analysis ................................................... 59
- 11 -
3.7- PMCA 2 Western blotting analysis .................................................................... 59 3.8- Western blotting with antiCalmodulin antibody................................................. 61 4- Expression of PMCA pumps in CHO cells ........................................................... 61
4.1- Continuos cell lines............................................................................................. 61 4.2- Transfection ........................................................................................................ 62 4.3- Isolation of microsomes from CHO cells ........................................................... 62 4.4- Mutagenesis ........................................................................................................ 63 5- ATPase activity...................................................................................................... 63 6- Immunocytochemistry ........................................................................................... 64 7- Computational methods ......................................................................................... 65 7.1- Sequence alignment ............................................................................................ 65 7.2- Structural model.................................................................................................. 65
RESULTS ................................................................................................................. 66
I. Expression and purification of PMCA 2 pump ................................................. 66
1- Expression of the PMCA 2 isoforms in Baculovirus system ................................ 67 1.1- Localization of PMCA 2 zb in insect cells ......................................................... 67 1.2- Expression and purification of PMCA2 zb ......................................................... 68 1.3- Time course of PMCA 2 zb expression in Sf9 cells ........................................... 68 1.4- Preparation of Sf 9 cells membranes for purification with a calmodulin resin ...................................................................................... 70 1.5- PMCA 2 zb solubilization and purification by a calmodulin affinity resin ................................................................................ 70 1.6- Construction of a Recombinant Baculovirus with a His-Tag Vector....................................................................................... 72 1.7- Time course PMCA 2 wa expression in Sf9 cells………………………………73
1.8- Sf 9 insect cells membrane preparation for His-Tag affinity purification ........................................................................ 74 1.9- PMCA 2 zb and wa purification by His-Tag affinity column..............................75 II. Biochemical study of mutants of plasma membrane calcium pump
isoform 2 causing deafness………………………………………………………..77 1- Construction and expression of PMCA 2 wa mutants……………………………78 1.2- Determination of the Ca2+ transport activities of microsomes from PMCA 2 zb, PMCA2 wa and of mutated pumps of the wa variant .................. 78 1.3- Study of a novel mutation near to the ATP binding domain of the PMCA 2 pump.................................................................................... 80 III. Functional properties of the PMCA 2 wa splice variant................................ 83
1- Functional studies of PMCA wa. Unusual variant of PMCA 2 pump.................................................................................................. 84 1.1- Expression of PMCA 2 splicing variants in CHO cells………………………...84 1.2- Determination of Ca2+ transport activities of Microsomes from CHO transfected CHO cells with PMCA zb, wb, wa and Tommy…………………..84 1.3- Calmodulin activation......................................................................................... 85 1.4- Activation by Phosphatidyl-serine...................................................................... 86
- 12 -
IV. Structural Model of PMCA 2 and biochemical study of the effect
of a Met265Ala mutation in one of the domain that mediate the
autoinhibition of the pump...................................................................................... 88
1- Comparative model between SERCA and PMCA 2 ............................................. 88 1.1-.Functional state of the PMCA 2 zb and M265A mutant..................................... 92
DISCUSSION ........................................................................................................... 96
REFERENCES....................................................................................................... 102
- 13 -
ABBREVIATIONS
AA Aminoacid
ATP Adenosine triphosphate
bp Base pair
BSA Bovine serum albumine
cDNA Complementary DNA
C12E8 Octaethyleneglycol mono-n-dodecylether
O.D. Optical density
DMEM Dulbecco’s modified Eagle’s medium
DMSO Dimethyl sulfoxide
DTT Dithiothreitol
EDTA Ethylene diamine tetraacetic acid
EGTA Ethylene glycol tetraacetic acid
ER Endoplasmic reticulum
FBS Fetal bovine serum
GAL Galactose
HEPES 4-(2-hydroxyethyl)-piperazineethanesulfonicacid
His (H) Histidine
IP3 Inositol 1,4,5-trisphosphate
IP3R Inositol 1,4,5-trisphosphate receptor
IPTG Isopropyl 1-thio-β-D-galactopyranoside
Kb Kilobase
kDa Kilodalton
LB Luria-Bertani
LDH Lactate Dehydrogenase
MOPS 3-(N-morpholino)propane-sulfonic acid
NADH Nicotinamide adenine dinucleotide
NCX Na+/Ca2+ exchanger
PAGE Polyacrylamide gel electrophoresis
PBS Phosphate-buffered saline
PEP Phosphoenolpyruvate
- 14 -
PK Pyruvate Kinase
PMCA Plasma membrane Ca2+ ATP-ase
PMSF Phenylmethylsulphonyl fluoride
PS Phosphatidyl Serine
PVDF Polyvinylidene Fluoride
RyR Ryanodine receptor
SR Sarcoplasmic reticulum
SERCA Sarco/endoplasmic reticulum Ca2+-ATPase
SDS Sodium dodecyl sulfate
Sf9 Spodoptera frugiperda insect cells
TAE Tris-acetate-EDTA
TE Tris-EDTA
Tris 2-amino-2-hydroxymethyl-1,3-propanediol
Triton X-100 Octyl phenol ethoxylate
TWEEN 20 Polyoxyethylene (20) sorbitan monolaurate
- 15 -
- 16 -
INTRODUCTION
The evolutionary transition from unicellular to multicellular life brought with it the
division of labor among cells. This in turn generated the necessity of exchanging signals
among the cells that formed the new organisms: as is well known, as a rule
monocellular organisms do not depend on the exchange of signals, their mutual
interactions being largely confined to the exchange of nutrients. Cell signaling is thus
essentially a hallmark of pluricellular organisms, in which a multiplicity of external
messages, carried by numerous first messengers, is translated inside the cells into
specific responses mediated by a less numerous group of newly produced chemicals, the
second messengers.
External signal, e.g. hormones, had long been known, but second messengers have only
arrived on the scientific scene more recently, with the discovery of cAMP in the 1950s.
One of the second messengers, however, had become known much earlier even if the
significance of the discovery was only appreciated much later, once the importance of
the observation became obvious: it was calcium (Ca 2+), which was the main actor in the
experiment published by Ringer in 1883 (1)
The advancement of knowledge rapidly became all-encompassing: as a large amounts
of new information was gathered, it gradually became clear that Ca 2+ had a number of
properties that made it unique among all other carriers of biological information. Three
of these properties are particularly striking. The first is the ability to act both as a first
and as a second messenger. The second is the autoregulative property of the Ca 2+
signal, i.e., Ca 2+ itself may control the generation and the regulation of the information
it carries. Finally, and perhaps most importantly, the Ca 2+ signal has a clear ambivalent
character: although essential to the correct functioning of cells, it can easily become a
messenger of death. ( Fig. 1)
Signaling operations typically involve the interaction of messenger chemicals with the
surface of target cells, which then process the incoming message and initiate an internal
signaling chain which is normally mediated by soluble signaling molecules. This chain
of events is the most common way of exchanging information between cells but is by no
means unique. Cells can also communicate with each other through direct contacts in
the form of gap junctions or through surface proteins that recognize protein counterparts
on the surface of other cells. As for first messengers, some may by pass the plasma
membrane and penetrate directly into target cells. Once there, they interact with
- 17 -
receptors in the cytoplasm and/or the nucleus and act without the intermediation of
second messengers. These alternative possibilities are all interesting and important:
however, the classical way of exchanging information remains that based on the
interaction of first messengers with plasma membrane receptors and on the generation
of diffusible second messengers inside cells. In considering Ca 2+ from this perspective,
a number of peculiarities emerges that are difficult to reconcile with the canonical
second messenger concept. Ca 2+ certainly behaves as a diffusible second messenger
generated (liberated) inside cells in response to the interaction of a number of first
messengers with plasma membrane receptors. However, the increase of cell Ca 2+ is not
directly linked to the processing of first messenger signals, as is the case, for instance,
for cAMP. It occurs one step further down the signaling chain and follows the
generation of another second messenger, e.g., InsP3, which will then liberate Ca 2+ from
intracellular stores. Strictly speaking, then, Ca 2+ should be defined as a third messenger.
But Ca 2+ can also penetrate directly into cells from the external spaces to initiate the
intracellular signaling chain. This, however, is not unique to Ca 2+: other second
messengers, e.g., the NO radical, may also have this dual origin (interstitial spaces and
intracellular ambient).
The peculiarities above may be considered minor and essentially formal. Another
distinctive property of the Ca 2+ signal, however, is more substantial. Ca 2+ has the
ability, now being documented in a growing number of cell types, to recognize specific
G-protein-linked plasma membrane receptors. Following the interaction, a G-protein-
mediated chain of events is initiated that ends in the modulation of the release of Ca 2+
itself from intracellular stores. The finding that Ca 2+ acted on a plasma membrane
receptor as any other canonical first messenger came as a surprise (2): the concentration
of first messengers is very low and necessarily fluctuates around cells, whereas that of
Ca 2+ is high and remains remarkably constant. Thus, the possibility that extracellular Ca
2+ would act as a first messenger had not been generally considered. However, it had
long been known that cells secreting the calciotropic hormones that regulate the
homeostasis of Ca 2+ in the organism, e.g., the cells of the parathyroid gland, modify the
release of the hormones in response to perturbations of external Ca 2+. Indications that
the mechanism by which these cells sensed the changes in external Ca 2+ may have
involved a plasma membrane receptor came recently. The first clear findings in this
direction were the discoveries that raising the level of external Ca 2+ induced a transient
- 18 -
and then a sustained elevation of Ca 2+ in parathyroid cells [12], which was linked to the
activation of phospholipase C and thus the generation of InsP3. (3)
1- General principles of Ca2+
signaling
Eukaryotic cells are surrounded by media containing free Ca 2+ concentrations that
exceed 1mM, but manage to maintain a free Ca 2+ concentration in the cytoplasm that is
four orders of magnitude lower. This very low internal concentration is dictated by the
necessity to avoid the precipitation of Ca 2+, given the poor solubility of Ca 2+-phosphate
salts. It also prevents the prohibitive expenditures of energy that would otherwise be
necessary to significantly change the background concentration of Ca 2+ in the vicinity
of the targets that must be regulated: if the resting free Ca 2+ in the cytosol were much
higher than nM, considerably larger amounts of energy would have to be invested to
change it substantially. The total Ca 2+ concentration within cells is naturally much
higher than nM but is reduced to the sub-µM ionic range first by binding to membrane
(acidic) phospholipids, to low molecular weight metabolites, to inorganic ions like
phosphate, and then by complexation to specific proteins.. These belong to several
classes: one comprises the membrane intrinsic proteins that operate as Ca2+ transporters
in the plasma membrane and in the membranes of the organelles. These proteins play
the most important role in the maintenance of the cellular Ca2+ homeostasis as they
move Ca2+ back and forth between the cytosol, where most of the targets of the Ca2+
signal are located, and the extracellular spaces or the lumenal spaces of the organelles.
Depending on whether large amounts of Ca2+ must be moved, or whether only the fine-
tuning of Ca2+ down to very low concentration levels is required, low or high Ca2+
affinity transporters become active.
The low-molecular compounds that have propensity to complex Ca2+ (e.g., phosphates,
ATP, aminoacids, acidic phospholipids of membranes) do so with low affinity, and also
easily complex other ions (e.g. Mg 2+) that may be more concentrated in the intracellular
environment. To achieve Ca2+ specificity, evolution has developed complex molecules
able to exploit the special chemistry of Ca2+ by offering to it specific binding sites. (4)
The ability of these sites to efficiently bind Ca2+ even in the greatly reduced ionic
concentration range resulting from the low affinity binding to the compounds
mentioning above, and to do so in the presence of much higher concentration of the
other cations, is made possible by the coordination flexibility of Ca2+.(5)
The
- 19 -
combination of charge and dimension of this ion makes it an ideal ligand for irregularly
shaped binding cavities, where the distance to the coordinating oxygens may vary
greatly. This could be convenient compared to Mg2+, whose ionic radius is much
smaller than that of Ca2+ and whose polarizing power much greater.
The conclusion that emerges is that the advantages of Ca2+ as a ligand for complex
(protein) molecules made it a particularly attractive candidate, perhaps even an obligate
one, as a messenger that would carry information to the intracellular ambient.
One final point must be mentioned before closing the discussion of the advantages of
Ca2+: The energetic metabolism of cells is phosphate-oriented: ATP is the universal
energetic currency, and it is self evident, given the propensity of Ca2+ phosphate salts to
precipitate, that high concentrations of the ATP cleavage product phosphate in the cell
ambient are only became possible if means were made available to keep the Ca2+
concentration very low. Evolution has evidently exploited the favorable properties of
Ca2+ as a ligand to develop means to efficiently lower its concentration to a point where
there would be no detrimental consequences for phosphate-oriented energetic
metabolism.
Figure 1: Ca2+
- modulated functions in mammalian cells
Ca2+
secretion
Ca2+ dependent transporters
gene expression
fertilization/ cell division
contraction
P
kinases/phosphatases
apoptosis
- 20 -
2- Ca2+
sensors EF- hand proteins are the most important and best-characterized Ca2+ sensors. They
bind Ca2+ using helix–loop–helix motifs (there are between 2 and 12 such motifs in the
various EF-hand proteins), which ligate Ca2+ to the side chain and carbonyl oxygen
atoms of some invariant residues. Variations in the coordination pattern have been
described. For example, oxygen atoms from neighboring helices can ligate Ca2+. Other
Ca2+ sensors — for example, the annexins, gelsolin and proteins that contain C2
domains (a compact -sandwich of two four-stranded -sheets) — should also be
mentioned, even if their role is not as general or as well understood as that of the EF-
hand proteins. Ca2+ sensors normally modulate target enzymes that are not Ca2+ sensors
themselves, but there are enzymes that can be modulated by Ca2+ directly, that is, they
are genuine Ca2+ sensors.
The founder member of the EF-hand protein family was the Ca2+-buffering protein
parvalbumin, as mentioned above. Its crystal structure was solved by Kretsinger and
Nockholds (6) in 1973, and its basic Ca2+-binding principles were extrapolated to other
proteins, the tertiary structures of which only became available much later. The family
of EF-hand proteins has now grown to include hundreds of members, the most
thoroughly investigated of which is calmodulin.
Most of what we now know about the processing of Ca2+ signals by EF-hand proteins
came from studying calmodulin. Ca2+ induces a conformational change that does not
affect the overall shape of calmodulin, but that makes its surface more hydrophobic
(Fig. 2). Calmodulin then approaches its targets, collapses around their binding domains
and transmits Ca2+ information to them. Ca2+-free calmodulin (apocalmodulin) can also
bind proteins through several different binding motifs (such as an IQ motif, which is a
consensus sequence of about 10 residues that begins with isoleucine-glutamine) and
may have an effect also in the absence of Ca2+. The proteins that are modulated by
apocalmodulin include enzymes (for example, phosphorylase kinase and some
unconventional myosins), structural proteins and membrane proteins (such as receptors
and ion channels) (7). As a modulator, calmodulin is 'uncommitted', that is, it interacts
with numerous targets and is associated with them reversibly. However, calmodulin can
also become permanently associated with some proteins; for example, this is the case
for the inducible form of nitric oxide synthase. Perhaps, calmodulin modulation is so
important to some enzymes that, once they have bound it, they never release it.
- 21 -
Figure 2: Conformational changes of Calmodulin (CaM) structure
Interestingly, enzymes have now been described that contain their own 'calmodulin',
either as a separate subunit, as is the case for the Ca2+-dependent protein phosphatase
calcineurin, or as a calmodulin-like domain that is integrated into the sequence. The
catalytic subunit of the ubiquitous isoforms of the protease calpain contains a carboxy-
terminal domain that is very similar to calmodulin. However, the example of calpains is
more complex and interesting, because their catalytic subunits also contain Ca2+-binding
sites other than EF-hand motifs (8).
Several other EF-hand proteins have become well known, for example, troponin C as
the Ca2+ receptor in myofibrils (9), recoverin as an 'actor' in the vision process, and
calbindin, the expression of which is controlled by vitamin D. Recently, calbindin has
been found to activate inositol monophosphatase, which is an important enzyme in the
phosphoinositide signalling pathway (10). Finally, a group of EF-hand proteins, which
are collectively called neuronal Ca2+ sensors, mediate neuronal functions such as the
release of neurotransmitters and the activity of a family of K+ channels.
•Apo-CaM, or CaM without calcium has a structure of two
globular domains connected by a flexible linker
•Each globular domain contains two EF-hand motifs (helix-
loop-helix).
•Each EF-hand motif contains two alpha helices connected by a
12 residue loop. The helices within each moitf are
perpendicular to one another.
•In apo-calmodulin the alpha helix of the two domains are
almost parallel to one another ("closed conformation").
The Structure of Apo-Calmodulin
•The calcium ions bind in the loop region of the EF-hand motif. When CaM
has bound calcium, its globular domains are perpendicular to one another
giving the "open conformation" (see blue/green).
•The C-terminal domain has higher affinity for Ca2+ and binds two
Ca2+ ions at all times. This is known as the low calcium state as only two of
the four possible binding sites are filled.
•The N-terminal domain with a lower affinity only binds Ca2+ ions when
necessary, for examples when CaM is regulating the amount of intracellular
calcium or binding to a target protein. This is the high calcium state due to
the occupation of all four binding sites.
The Structure of Calmodulin with Ca2+
- 22 -
3- Ca2+
-transporting proteins
Membrane-intrinsic proteins can create and shape Ca2+ signals by functioning as
channels, ATPases (pumps), or exchangers that transport Ca2+ across membranes. Ca2+
channels mediate the penetration of Ca2+ into the cell and its exit from the
sarco(endo)plasmic reticulum. In the plasma membrane, Ca2+ channels are gated by
voltage changes, by ligand interactions, or by a mechanism that is linked to the
emptying of intracellular Ca2+ stores and that has become known molecularly only
recently (11). The voltage-gated channels have several sub-types (12), of which the L-
type, which is the target of widely-used Ca2+ antagonists (13), is the best characterized.
The most important ligand-operated channels are gated by neurotransmitters and,
among them, the glutamate-operated channels are the most intensively studied (14). In
the sarco(endo)plasmic reticulum, the opening of Ca2+ channels is regulated by Ca2+
itself — the process of Ca2+-induced Ca2+ release(15) — but this opening also requires
ligands like inositol-(1,4,5)-trisphosphate (Ins(1,4,5)P3) (16) or, in selected cell types,
another endogenous ligand, cyclic ADP ribose (17), which releases Ca2+ through
channels that are known as ryanodine receptors. Isoform 1 of this receptor (RYR1) is
expressed predominantly in skeletal muscles, whereas isoform 2 (RYR2) is expressed in
the heart and brain. The link between the activation of plasma-membrane receptors and
the production of Ins(1,4,5)P3 is well documented, but data on the link between plasma-
membrane receptors and the production of cyclic ADP ribose are scarce. In addition,
cyclic ADP ribose is thought to be produced extracellularly by an ectoenzyme (18) and
therefore must be imported into cells to function. Interestingly, free ADP ribose — a
product of NAD+ hydrolysis and a breakdown product of cyclic ADP ribose — that is
produced inside cells has been shown to gate a Ca2+-entry channel, which is known as
long transient receptor potential channel 2 (LTRPC2) (19). Finally, the nicotinic acid
derivative of NADP+ — NAADP+ — is the latest addition to the family of Ca2+-
mobilizing messengers (17)
Both the plasma membrane and the inner mitochondrial membrane contain Na+/Ca2+
exchangers (NCX and MNCX, respectively) that export Ca2+ from the cell (20,21) and
the mitochondrial matrix (22), respectively. The high capacity plasma-membrane NCX
is particularly active in excitable cells, which demand the periodic ejection of large Ca2+
loads. The less well-characterized mitochondrial exchanger (MNCX) is also particularly
- 23 -
active in excitable cells. Whereas the plasma-membrane NCX exchanges three Na+ per
one Ca2+, and therefore responds both to the Na+ and Ca2+ transmembrane gradients and
to the voltage difference across the plasma membrane, MNCX seems to exchange two
Na+ per one Ca2+. This electrically neutral operation of MNCX allows the export of
Ca2+ from mitochondria in the presence of the large, negative membrane potentials that
are maintained at the matrix side of the inner mitochondrial membrane by the
respiratory chain (Fig. 3).
Figure 3: The Ca
2+ transporters
of cell membranes: PMCA: Plasma Membrane Ca2+
Pump; NCX: Na+/Ca2+exchanger; SERCA Sarco/Endoplasmic Ca2+ Pump; Insp3R: Inositol-1,4,5 Trisphosphate Receptor; RyR: Ryanodine Receptor
A spectacular advance in the field of membrane Ca2+ transporters was the determination
of the tertiary structure of the sarco(endo)plasmic reticulum Ca2+ ATPase (the SERCA
pump) by Toyoshima et al. (23,24). The structure was solved both in the Ca2+-bound
(E1) and Ca2+-free (E2) conformations, which validated predictions of the membrane
topology and molecular mechanism of the pump. The SERCA1 isoform is expressed
predominantly in fast-twitch muscles, whereas SERCA2 is expressed in numerous non-
muscle tissues as well. Splicing isoforms have also been described. One important
aspect of the SERCA-pump function is the regulation by the membrane intrinsic protein
phospholamban, the amino-terminal hydrophilic domain of which interacts with the
active centre of the pump to keep it inactive (25). Phosphorylation of two neighbouring
Endo(sarco)plasmic
reticulum
Nucleus
Golgi
Ca (10 M)2+
++
++
---
Mitochon-
drion
-
-3
InsP3 / cADPr
(Ca2+)
solubleCa2+ binding proteins
ATP
ADP
Ca
2+
InsP3
Ca 2+
ATP
ADP
Ca2+
Ca 2+
Na+
ADPATP
Na+
Na+
Ca ?2+
Ca 2+ (10 -7 M)
Endo(sarco)plasmic
reticulum
Nucleus
Golgi
Ca (10 M)2+
++
++
---
Mitochon-
drion
-
-3
InsP3 / cADPr
(Ca2+)
solubleCa2+ binding proteins
ATP
ADP
Ca
2+
InsP3
Ca 2+
ATP
ADP
Ca2+
Ca 2+
Na+
ADPATP
Na+
Na+
Ca ?2+
Ca 2+ (10 -7 M)
- 24 -
residues in phospholamban detaches it from the pump and relieves the inhibition (Fig.
4)
Figure 4: A)Architecture of the sarcoplasmic reticulum Ca2+-ATPase. Toyoshima, C. et al. Crystal structure of the calcium pump of sarcoplasmic reticulum at 2.6 Å resolution. Nature 405, 647–655 (2000) B) PLB/SERCA proposed interaction.
MacLennan D. and Kranias G. Phospholamban: a crucial regulator of cardiac contractility. Nature Reviews Molecular Cell Biology 4, 566-577 (2003)
Besides the well-known sarco/endoplasmic-reticulum Ca2+-transport ATPases
(SERCA), animal cells contain a much less characterized P-type Ca2+-transport ATPase:
A
B
- 25 -
the PMR1/SPCA Ca2+/Mn2+-transport ATPase. SPCA is targeted to the Golgi apparatus.
Phylogenetic analysis indicates that it might be more closely related to a putative
ancestral Ca2+ pump than SERCA. SPCA supplies the Golgi apparatus, and possibly
other more distal compartments of the secretory pathway, with the Ca2+ and Mn2+
necessary for the production and processing of secretory proteins. In the lactating
mammary gland, SPCA appears to be the primary pump responsible for supplementing
the milk with high (60-100 mM) Ca2+. It could also play a role in detoxification of cells
overloaded with Mn2+. Mutations in the human gene encoding the SPCA pump
(ATP2C1) result in Hailey-Hailey disease (Fig 5), a keratinocyte disorder characterized
by incomplete cell adhesion. Recent observations show that the Golgi apparatus can
function as a Ca2+ store, which can be involved in setting up cytosolic Ca2+ oscillations.
(26, 27)
Figure 5: Hailey-Hailey disease (HDD) mutations in human Golgi/secretory
pathway Ca2+
,Mn2+
-ATPase (hSPCA1). Ton V K, Rao R. Functional expression of heterologous proteins in yeast: insights into Ca2+ signaling and Ca2+-transporting ATPases. Am J Physiol Cell Physiol. 287(3):C580-9. (2004).
- 26 -
4- THE PLASMA MEMBRANE CALCIUM PUMP (PMCA)
The PMCA pump was discovered by H. J. Schatzmann forty years ago (28) as a system
responsible for the active extrusion of Ca2+ from red cells. Later on, Schatzmann and
Vincenzi (29) demonstrated that Ca2+, when transported, activated the hydrolysis of
ATP at the internal surface of the membrane. Following the initial observations on the
red cell membrane a large number of studies on numerous cell types characterized the
ATPase as a bona fide member of the family of transport ATPases that where later
classified as P-type pumps (30,31). These pumps are characterized by the formation of
an energized intermediate in which the -phosphate of ATP is transferred to an
invariant Asp in the active site of the enzyme, resulting in the formation of a high-
energy acyl-phosphate. As in all other P-type pumps, the reaction cycle of the calcium
pumps oscillates between two conformational states: the E1 state, in which the enzymes
have high affinity for Ca2+, and the E2 state, in which the enzymes have much lower
Ca2+affinity. Ca2+ is bound to the cytosolic domain of the enzyme in the high-affinity
conformation (E1), prior to the phosphorylation of the catalytic aspartic acid by ATP. A
conformational change of the phosphorylated enzyme (E1P→E2P) then exposes bound
Ca2+ to the extracellular/lumenal side of the membrane and promotes its release prior to
the cleavage of the phosphorylated intermediate, after that, the pump returns to the E1
conformation ( Fig. 6).
Figure 6: The P-type pumps catalytic cycle
E2
Ca2+
-P
Cytosol
-P E2
Ca2+
E2
P
E1
Ca2+
~P E1
ATP
Ca2+
E1
ATP
Ca2+
AD
ATP
P
Extracellular space
- 27 -
Three P-type Ca2+ pumps are now recognized: The Golgi-membrane SPCA pump is the
most recent entry in the field and has not yet been studied as extensively as the other
pumps: thus, most of the discussion on general properties in the next paragraphs will
concentrate on the pumps of the plasmamembrane (PMCA) and of the
endo(sarco)plasmic reticulum (SERCA), for which considerable information is
available. The two pumps share basic properties, but also have significant differences.
For instance, the SERCA pump transports two Ca2+ ions per molecule of ATP
hydrolyzed, whereas the Ca2+ /ATP stoichiometry of the PMCA pump is 1:1. Both
pumps exchange Ca2+ for H+: early reports on the reconstituted PMCA pump had
indicated an electroneutral Ca2+/H+ exchanger (32), but more recently work has instead
convincingly shown partial electrogenicity, i.e. one H+ is exchanged for one Ca2+ (33).
The single Ca2+ binding site of the PMCA pump, which is thought to correspond to site
II in the SERCA pump, is likely to be located on the cytosolic site of the membrane
prior to the translocation step, and to face the exterior of the cell at its end. After
releasing Ca2+ to the exterior of the cell, the pump becomes dephosphorylated and
eventually returns to the E1 conformation. The process is fully reversible; the
phosphorylated intermediate can be formed, under appropriate conditions, from
inorganic phosphate as well.
As with all other P-type pumps, the PMCA pump is rapidly and irreversibly inhibited by
micromolar concentrations of the phosphate analogue orthovanadate, which binds to the
E2 conformer of the pump, thereby preventing the E2/E1 transition (34). The other
general inhibitors of the P-type pumps are lanthanide ions, which have been shown to
displace Ca2+ by combining with its high-affinity binding sites on the SERCA pump
(35, 36). Since they allow phosphorylation to occur, they could replace Ca2+ as the metal
ion that activates the phosphorylation as suggested for the SERCA pump (36,37). As
expected, lanthanides also inhibit the PMCA pump, however, at variance with the
SERCA pump (and other P-type pumps), they markedly increase the steady-state level
of the phosphoenzyme (38). The effect is attributed to the inhibition of the hydrolysis of
the phosphorylated intermediate by La3+: it is very advantageous, as it permits the
recognition of the PMCA pump in autoradiographic gels of membrane preparations
containing much higher amounts of the SERCA or other P-type pumps.
The peptide caloxin 2A1 has been recently reported to inhibit the PMCA pump (39) by
binding to putative extracellular domain 2 of isoform 1 (see below for information on
the membrane topology of the pump) which is one of the two PMCA isoforms of
- 28 -
erythrocytes. Another component of the caloxin’s peptide family, caloxin 1A1,
apparently blocks the pump by perturbing its first extracellular domain. The caloxin
findings suggest that transmembranes domains 1 and 2 play a role in the reaction cycle
of the PMCA pump (40).
The PMCA pump is essential for the control of the cytosolic Ca2+ concentration in all
eukaryotic nonmuscle cells, but in muscle cells, particularly heart cells, is likely to be of
minor importance with respect to the much more abundant SERCA pump and to the
high transport capacity sodium calcium exchanger of the plasma membrane. In
eukaryotic cells, the amount of PMCA pump is thought not to exceed the level of 0.1%
of the total membrane protein. This value may perhaps be underestimated in neurons,
where the level of expression of the pump is apparently higher than in the other cells
types (41).
4.1- ISOLATION AND PURIFICATION OF THE CALCIUM PUMP
The purification of the PMCA pump was first reported in 1979 by Niggli et al. (42)
using calmodulin affinity chromatography. The decision to use calmodulin column was
prompted by previous work on erythrocyte membranes that had shown that calmodulin
activated the pump (43, 44). Red blood cell membranes were solubilized with Triton X-
100 and stabilized with phosphatidylserine and the preparation was applied to the
column. Most of the membrane proteins, including the Mg2+-dependent ATPase
activity, were eliminated with Ca2+ containing buffers. EDTA then eluted from the
column the Ca2+ dependent ATPase as a major band of about 125 KDa accompanied by
a minor band of about 205 KDa. Both bands were phosphorylated by ATP in the
presence of Ca2+ and Mg2+: the heavier band was later shown to be a dimer of ATPase.
The PMCA was then successfully reconstituted as an active pump in liposomes. A
crucial advance that resulted from the work on liposomes was the detailed analysis of
the role of acidic phospholipids as activators. In the first purification experiments the
pump had already been maximally active in the absence of calmodulin. This was so
because the calmodulin column had been saturated with the acidic phospholipid
phosphatidylserine, which was later found to activate the pump alternatively to
calmodulin (see next section). Replacement of phosphatidylserine with the zwitterionic
phospholipid phosphatidylcholine soon produced purified preparations that were fully
responsive to calmodulin. Another important result of the early work on the purified
- 29 -
PMCA was the precise definition of the pattern of proteolysis, especially of that by
calpain (45). It occurred in two steps: the first calpain cleavage removed about half of
the calmodulin binding domain and the portion of the pump downstream of it. A second
cut then truncated the pump at the beginning of the calmodulin-binding domain. The
fully active 124 kDa peptide resulting from the two calpain cuts could be easily purified
since it was not retained by calmodulin columns, and was later used to demonstrate that
the calmodulin binding sequence acted as an autoinhibitory domain. It bound to two
sites in the cytosolic domain of the pump close to the active centre (see next section),
maintaining the pump inactive (46, 47). The irreversible activation of the pump by
calpain was thus evidently due to the removal of the autoinhibitory region, i.e., the
calmodulin-binding domain. It may have significance in pathological states of the cell,
when the resting level of free Ca2+ remains elevated for long periods, thus activating
calpain (48).
4.2- CLONING OF THE PUMP AND RECOGNITION OF ISOFORMS
The primary structure of the pump was deduced in 1988 using rat brain (49) and human
teratoma (50) cDNA libraries, respectively. Two rat sequences contained 1176 (129.5
KDa) and 1198 (132.6KDa) amino acids, respectively, the human sequence 1220
residues (137.7KDa) corresponding to that of the latter rat isoform. All regions of
functional importance present in other ion-transporting ATPases were soon identified.
The translated sequence had the calmodulin-binding in the carboxyl terminal cytosolic
tail. It also had a domain near the amino terminus very rich in basic aminoacids which
was soon found to have a role in the interaction with activatory acidic phospholipids
(see below).
It soon emerged that several PMCA isoforms were encoded by a multigene family, and
that additional isoform variability was produced by alternative RNA splicing of each
gene transcript.
All regions important to the catalytic function of P-type pumps were found to be highly
conserved in the PMCA sequences and, as expected, no isoform diversity was detected
in these regions, which are essential to the catalytic function of the pump. The structural
motifs responsible for the membrane-folding pattern of the P-type pumps also tend to be
conserved, i.e., isoform diversity was not found to involve the general arrangement of
the transmembrane domains of the pump.
- 30 -
4.3- THE PLASMA MEMBRANE Ca2+
PUMPS: STRUCTURAL AND
REGULATORY CHARACTERISTICS.
The cloning work predicted the PMCAs to contain 10 membrane-spanning segments
(TM), with the NH2 and COOH termini located on the cytosolic side of the membrane.
The bulk of the protein mass faces the cytosol and consists of three principal domains:
the intracellular loop between TM 2 and 3, a large unit between TM domains 4 and 5,
and the extended tail following the 10th transmembrane domain. The first intracellular
loop between membrane spanning domains 2 and 3 corresponds to the transduction or
actuator domain of the SERCA pump (51), and is thought to play a role in the long
range transmission of conformational changes occurring during the transport cycle. The
large cytosolic region (about 400 residues) between transmembrane segments 4 and 5
contains the catalytic domain (the ATP binding site and the invariant aspartate that
forms the acyl phosphate intermediate). Finally, the extended COOH-terminal tail is the
major regulatory domain of the PMCAs. On the basis of sequence comparison and
computer modeling (unpublished), the overall 3-D structure of the PMCAs resembles
that of the SERCA pump. However, differences between the two types of calcium
pumps are also evident: the COOH-terminal tail is much shorter in the SERCA than in
the PMCA pumps, and the PMCAs contain a basic sequence (see above) in the
transduction domain that binds regulatory acidic phospholipids (52).
Long before the first PMCA cDNA was cloned, it was realized that the pump was
tightly regulated by Ca2+ bound to the modulator protein calmodulin (see above), which
decreased the Kd of the pump for Ca2+ from 10- 20 µM in the resting state to less than 1
µM: calmodulin interacts with the C-terminal tail of the pump (53). In the absence of
calmodulin, and as long as the Ca2+ concentration in the vicinity of the pumps is low (<
50−100 nM), the PMCAs remain in an auto-inhibited state. This is so because the
calmodulin binding domain in the C-terminal tail makes intramolecular contacts with
the first and second cytosolic loop of the pump, thus presumably preventing substrate
access to the active site (46, 47). Ca2+ loaded calmodulin de-inhibits the pump by
removing its C-terminal tail from the intramolecular binding sites, increasing Ca2+
affinity and raising Vmax. (54).This regulatory mechanism is similar to that of other Ca2+-
calmodulin dependent enzymes. In the case of one of these enzymes, calmodulin
dependent protein kinase I (CaMKI), the crystal structure has convincingly shown that
in the absence of Ca2+-calmodulin the COOH-terminal regulatory domain (which
- 31 -
overlaps with the calmodulin binding domain.) interacts with multiple sites in the
catalytic core, preventing substrate access and obstructing ATP binding (55) exactly as
predicted for the case of the PMCA pump. One additional comment is in order on the
interplay between the PMCA pump and calmodulin: unique among calmodulin targets,
the pump can be activated by the C-terminal half of calmodulin. Since the latter has
peculiarly high Ca2+ affinity, it is probably Ca2+-saturated even at the low level of Ca2+
in the resting cell. It could thus interact with the N-terminal half of the calmodulin
binding domain, maintaining the pump in a low state of activity even under conditions
of resting cytosolic Ca2+ (56).
In addition to calmodulin, kinase mediated phosphorylations also activate the pump.
The effect was first identified in the pump of heart sarcolemma membranes and
attributed to the cAMP protein kinase (PKA) (Fig 7). The phosphorylation step
increased both the Vmax and the Ca2+ affinity of the pump, decreasing the Km (Ca2+) to
about 1 M in the absence of calmodulin. The consensus site for PKA (KRNSS) was
identified downstream of the calmodulin-binding domain in the C-terminal region of
one of the isoforms of the pump (PMCA1) (57).
Figure 7: Release of Ca-ATPase Inhibition. Motional restriction of catalytically important motions of the nucleotide binding domain (N) of the Ca-ATPase occurs through the binding of either PLB (A) or an autoinhibitory domain (shown in red), which underlies the regulation of calcium transport in the sarcoplasmic reticulum or plasma membrane of the heart. In both cases, the stabilization of helical structure within a conformational switch region results in a shortening of the physical dimensions and the necessary release of the inhibitory interactions with the N-domain.
- 32 -
The problem of activation by protein kinase C (PKC) is more complex. Stimulation of
the pump by this kinase has been described by several authors (58-64), also in the
solubilized, purified erythrocyte Ca2+ pump (65). The phosphorylation occurs in the C-
terminal region, involving both serine and threonine residues, one of the
phosphorylation sites being a threonine residue within the calmodulin binding domain.
Based on peptide data (66), phosphorylation would be expected to prevent binding of
calmodulin to the pump while causing a calmodulin-like activation (67).
As mentioned above, PMCA pump stimulation is also induced by acidic phospholipids
and polyunsaturated fatty acids. The observation that acidic phospholipids are more
effective than calmodulin, i.e. that they decrease the Km (Ca2+) to values lower than
these induced by calmodulin (68), has led to the suggestion that acidic phospholipids
and calmodulin activate the Ca2+ ATPase by separate mechanisms involving different
binding sites. Work with distinct proteolytic fragments of the PMCA pump and with
synthetic peptides has indeed revealed two separate phospholipid binding regions in the
pump: one corresponding to the calmodulin binding domain, which has basic character
(69), and the other, as mentioned, situated in the first cytosolic loop (52). The
suggestion that a substantial portion of the phospholipid stimulation of the pump was
apparently mediated by the calmodulin binding domain is in line with the observation
that also other calmodulin-regulated proteins, i.e., neuromodulin and myosin light-chain
kinase, are activated by acidic phospholipids.
The phospholipid effect could be important in vivo, as the ATPase in the membrane is
presumably surrounded by an amount of phosphatidylserine adequate for about 50% of
maximal activation (70). However, the concentration of phosphatidylserine, and of the
other abundant acidic phospholipids in the plasma membrane, is presumably invariant.
By contrast, the concentration of phosphatidylinositol (PI) and of its two
phosphorylated products (PIP1 and PIP2) in the membrane is modulated in response to
first messenger challenges to the cell, making them more attractive as potential
modulators. Thus, an activation/deactivation cycle of the pump linked to the metabolism
of PI and designed to modulate the export of Ca2+ from the cell as demanded by stimuli
that promote intracellular Ca2+ release, appears possible (71).
In addition to mediating the regulation by Ca2+-calmodulin, the C-terminal region of the
PMCAs has also been shown to be affected by the irreversible activator calpain (see
above 45, 72, 73), and to mediate the interaction with PDZ domain proteins. The
interaction may be important for the local organization of Ca2+ signaling domains at the
- 33 -
plasma membrane and/or for anchoring Ca2+ regulatory complexes to the cytoskeleton.
The PMCAs are also activated by self-association (74). Even if self-association still
permits to bind Ca2+-calmodulin (75), in the oligomerized state there is no further
activation by calmodulin. The functional significance of the dimerization
(oligomerization) of the PMCAs, which is mediated by the calmodulin binding domain
(76), is a controversial point. An increase of Ca2+-ATPase activity at high enzyme
concentrations has been attributed to pump oligomerization (75). However, more recent
findings (77), suggest that the dimerization does not correlate with the function of the
pump, or has minimal effects on it (78).
4.4- ISOFORMS OF THE PMCA PUMP
As mentioned, four basic PMCA isoforms are encoded by four independent genes
named ATP2B1 –ATP2B4, that map respectively to the following human chromosomal
loci: 12q21-q23, 3p25-p26, Xq28 and 1q25-q32 (79-82).
The size of the four human genes ranges from about 60kb (PMCA3) to over 350 kb
(PMCA2) (83-86), however, the number and size of protein-coding exons are highly
conserved among the 4 genes, suggesting a common origin by duplication of an
ancestor gene. In all genes the first exon encodes a 5’ prime untranslated sequence that
precedes the ATG starting codon in the second exon. The transcribed proteins have a
homology of about 70-80 % and maintain the general folding features. Residues differ
among isoforms especially in the N- terminal protrusion and the C-terminal tail.
PMCA1 and PMCA4 are ubiquitously distributed in animal cells and are presumed to
carry out housekeeping functions to maintain Ca2+ homeostasis. PMCA1 appears to be
ubiquitously expressed since the earliest states of development, as highlighted in studies
of mice pump phenotypes. PMCA4 is widely expressed in human tissues and supports
PMCA1 in maintaining Ca2+ homeostasis. In human cells PMCA4 is particularly
abundant in kidney, erythrocytes and testis, where it is the principal isoform. PMCA2
and 3 are tissue restricted, and found in particular in the nervous system, suggesting
functional specificity.
In the nervous system PMCA2 and PMCA3 are present with regional and
developmental distribution differences. PMCA2 is abundant in the Purkinje cells of
cerebellum and in the outer hair cells of the inner ear (87, 88), while PMCA3 is
abundant in the habenula and in the choroid plexuses (89, 90). However, traces of
- 34 -
PMCA3 are also present in skeletal muscle (91). Hippocampus cells, instead, are
particularly rich in PMCA1, suggesting an additional tissue specific function of this
ubiquitous isoform. Outside the nervous system, PMCA2 is also abundantly expressed
in lactating mammary glands, and significant levels of this isoform have also been
found in the kidney. During development, isoforms 2 and 3 appear at about day 12.5 of
gestation and vary in levels and time distribution: PMCA2 starts to be detected in high
levels in the dorsal root ganglia and in the retinoblasts of the developing eye. PMCA3
could be important for the correct development of the organs, as it is widely expressed
in tissues from day 12.5 up to day 16.5 of gestation, when it start to be selectively
expressed in cell types (92).
Information on the differential functional properties of the 4 basic PMCA isoforms is
scarce. The calmodulin affinity of the two tissue-specific variants, PMCA2 and
PMCA3, is much higher than that of the two ubiquitously expressed isoforms PMCA 1
and PMCA 4 (41), and the sensitivity to calpain is particularly elevated in PMCA1 (93).
When tested in the native intracellular environment the two recombinant ubiquitous
isoforms are far less effective in controlling temporary Ca2+ increases induced by the
challenging of model cells overexpressing the pumps with agonists that produce Ca2+
transients, than the two tissue specific isoform PMCA2 and PMCA3 (94).
4.5- SPLICING VARIANTS
RNAs encoding for plasma membrane calcium pumps are subjected to alternative
splicing (95) at two different sites, considerably increasing the number of pumps
variants. About twenty of the many putative variants have been identified, some,
however, only at the RNA level. The physiological rationale for the existence of such a
great number of tissue specific variants is one of the open problems in the PMCA pump
area.
Supplementary exons are alternatively transcribed in the proximity of the two main
regulatory domains of PMCAs. Splice site A is located upstream of the phospholipid
binding domain in the first intracellular loop of the pump, and splice site C is placed
inside the calmodulin binding domain in the C terminal cytosolic tail of the protein. A
third splice site, Splicing site B, is now considered an aberrant splice variant.
Alternative splices at sites A and C have been extensively characterized for human and
rat isoforms and reviewed in numerous papers (96-97). A possible additional variant has
- 35 -
so far only been identified in amphibians (98). The scheme of Fig. 8 summarizes the
variants so far identified for each isoform.
Figure 8: Membrane Topology domains and splicing variants of the PMCA pump
Two different nomenclature systems (TABLE 1) have been developed to identify the
splicing variants (99, 100). According to the most commonly used one, insertions in
splice A are addressed with letters from z to w, according to the increased number of
inserted exons: (for instance, in variant w of PMCA2 three exons are included, in
variant x only one exon of 42 bp is transcribed, in variant y two exons, one of 33 and the
other of 60bp are inserted, in variant z all supplementary exons are excluded). At splice
site C the alternatively transcribed exons are identified by letters from a to e. At
variance with site A, at splice site C alternative exons are fully or partially transcribed
determining changes of the reading frame and the premature truncation of the protein.
- 36 -
Original Nomenclature
Alternative
Nomenclature
Splice site A z AI
x AII
y AIV
w AIII
v AV
Splice site B a CII
b CI
c CIII
d CIV
e CV
f CIV
PMCA1 is characterized by the presence in all identified mature transcripts of the
sequence of an exon of 40 bp, which is only alternatively included in the other isoforms,
therefore all known PMCA 1s correspond to the x variant. PMCA 2 presents a more
complex pattern of possible splice variants for site A. Up to three different exons
(respectively of 33, 60 and 42 bp) can be transcribed. So far only 4 of the 8 possible
combinations have been identified. Variant y has been so far detected in rat but not in
humans. Variant v, in which a fourth transcribed exon is inserted at splice site A, has
been found only in bullfrog. PMCA 3 and 4 exist only as variants x or z, including or
excluding an exon of 42 or 36 nucleotides, respectively.
Splicing at site C occurs with different degrees of complexity for each isoform, however
common options can be identified. A multitude of variants is generated by exon-internal
splice sites, read-through into adjoining introns, and inclusion or omission of entire
- 37 -
exons. As mentioned, such a combination of events usually leads to reading frame
changes with the consequent insertion of premature stop codons, and modification of
the C terminal structure of the pump. The most common modification determined by
total or partial insertion of exons is the truncation of the calmodulin binding domain,
with changes in the regulatory mechanism of the pumps. Unmodified (no inserts) C-
terminal proteins are designated as b variants. In general, for PMCA1 and PMCA 4 a
single exon is involved in alternative splices at site C, while PMCA2 and PMCA3
include or exclude partially or completely 2 exons.
The tissue distribution and the targeting to plasma membrane domains, of the splice
isoforms has not been determined in detail, but it is known that the ubiquitous variants
PMCA1 xb and PMCA4 b localize predominantly in the basolateral part of the plasma
membranes of polarized cells. PMCA 2 wb is expressed in mammary gland cells and
localizes instead in the apical section of the cell membrane. The w insert probably has a
general role in the membrane sorting of PMCA variants: in polarized cell lines the
splicing w is a determinant for the targeting of the pumps to the apical plasma
membrane (101). Interestingly, PMCA 2 wa is selectively expressed in cochlear and
vestibular stereocilia, as shown by immunoprecipitation from purified hair bundles of
rat utricles (102). Recently, it has also been demonstrated that the combination of splice
w at site A and a at site C is essential for the correct targeting of the protein to the
stereocilia (103).
4.6- PROTEIN INTERACTORS OF PMCA PUMPS
Naturally, the established function of PMCAs is to expel Ca2+ from cytosol. But an
additional role for the pump has been recently suggested, on the basis of newly
discovered interactions of the pumps with molecules involved in signaling pathways
(104, 105): the pump would function as a modulator of signal transduction. For
example, in cardiomyocites, where the Na/Ca2+ exchanger is far more important than
the PMCA pump in the extrusion of Ca2+, PMCA4 has been shown to form a ternary
complex with α1-syntropin and nitric oxide synthase1 (NOS-1). The physical
interaction with NOS-1 induces a negative regulation of the NO production, and the
overexpression of α1-syntropin synergistically increases the inhibitory effect. NOS-1
binds to the PDZ binding domain in the C-terminal tail of PMCA, while the α1-
- 38 -
syntropin interacts with the distal region of the large intracellular loop of the protein
(residues 652-840) (106).
The same region of the main intracellular loop of PMCA4 binds the tumor suppressor
Ras-associated Factor 1 (RASSF1), a Ras effector protein involved in H-Ras-mediated
apoptosis. Co-expression of RASSF1 and PMCA4b inhibits the epidermal-growth-
factor dependent activation of Ras (107).
Another interactor of the pump that binds to its second intracellular loop (residues 501-
575) is calcineurin, whose modulation of the transcriptional activity of NFAT is
inhibited by the interaction. PMCA would therefore be an upstream effector of
calcineurin/NFAT signaling, acting as a regulator of differentiation and adaptation of
different tissues (108). It would seem logical to think that the interaction of the pump
with partners should be isoform-specific, and the search of isoform-specific interactors
has recently been succesful. The 14-3-3 protein has been found to interact with PMCA4
but not with PMCA2, when coexpressed in model cells. The interaction occurs at the N-
terminal portion of the pump and is inhibitory (109).
Interactors of the PMCA pumps were also identified using the C-terminal tail of the
pump, which contains a PDZ binding sequence, as bait in yeast double screening
experiments. These include members of the membrane associated guanylate-kinase
family (MAGUK), cytoskeletal proteins, the Na+/H+ exchanger regulatory factor 2
(NHERF2), the Ca2+/Calmodulin- dependent serine protein kinase (CASK) and the
novel protein PMCA-interacting single-PDZ protein (PISP). CASK co-precipitates with
PMCA4 from brain and kidney tissues lysates. The CASK protein is a co-activator of
promoters containing transcription of T-elements. Overexpression of PMCA
downregulates T-element dependent reporter activity (105).
As for PISP, it transiently interacts with PMCA b variants and plays a role in the sorting
of the pump to or from the plasma membrane (110).
PMCA4b interacts with the PDZ domains of several MAGUKs, a result that has led to
the suggestion that C-terminal splicing would generate pump isoforms that may be
differentially recruited to multifunctional protein complexes involved in Ca2+ regulation
(111).
PMCA2b and PMCA4b have been used to explore the association with NHERF1 and
NHERF2. The interaction with the latter was found to be specific for PMCA2b, i.e., it
did not occur with PMCA4b. PMCA2b apparently preferred NHERF2 over NHERF1 as
an interactor. The interaction may provide a link between the Ca2+ pump and the actin
- 39 -
cytoskeletal network, potentially stabilizing the pump in a particular membrane
microdomain, possibly allowing its regulation by co-assembled protein kinase A.
Alternatively, NHERF2 mediated co-clustering of multiple PMCA 2b molecules may
facilitate their oligomerization, which has been shown in a previous section to lead to
calmodulin-independent pump activation (112).
One more example of an attractive partner protein for b variants of all PMCAs via their
PDZ binding domain is the Ania 3/Homer protein. The Homer family of proteins has
gained attention as a component of the post-synaptic density and is involved in the
coupling of NMDA glutamate receptors to the emptying of Ca2+ stores. Coexpressed
Ania 3 colocalized with PMCA at the plasma membrane of polarized MDCK epithelial
cells, and endogenous Ania3/Homer and PMCA2 co-expressed in the soma and
dendrites of primary hippocampal neurons (113). Fig. 9 offers a comprehensive view of
the partners of PMCA pumps identified so far and of their approximate sites of
interaction.
Figure 9: Interactors of Plasma Membrane Ca2+ ATPase pumps
- 40 -
4.7- PMCA PUMP AND DISEASE
4.7.1- PMCA PUMP KNOCKOUT MICE
Knock out mice have been developed and the phenotypes extensively studied for all
isoforms of the PMCA pumps with the exception of PMCA 3. As this isoform is widely
expressed in tissues of developing embryos its role may be critical for normal
development of gestation. Hopefully, PMCA 3 gene ablation studies will be soon
forthcoming.
PMCA1 was disrupted by targeting the catalytic phosphorylation site, but homozygous
knockouts resulted in embryonic lethality. Null mutant embryos were identified up to
day 3 of gestation, but not during the period of organogenesis. The inability to generate
adult live animals underlines the important role of this housekeeping isoform from the
earliest ages of development. On the other hand, heterozygous mutants did not to
present a pathological phenotype, even if the smooth muscle of blood vessels (the portal
vein was examined) appeared apoptotic. As this smooth muscle does not express the
other ubiquitous isoform PMCA 4, the absence of PMCA 1 gene on one allele was
evidently inefficiently compensated (114).
Mice subjected to the targeted ablation of the PMCA 4 survived and appeared
outwardly healthy. Histological analysis of organs provided no evidence of major tissue
alterations or in vivo cell death (115). Thus, in spite of its ubiquitous expression, PMCA
4 appears to be less critical than PMCA 1 in the maintenance Ca2+ homeostasis. A major
phenotypic alteration was nonetheless detected, and this was male infertility. Sperm was
unable to achieve efficient hyperactivated motility, and was unable to reach and fertilize
the egg. This was evidently due to the fact that isoform 4 represents 90% of all PMCA
pumps expressed in testis cells and underlies the importance of PMCA in the regulation
of Ca2+ homeostasis in sperm cells, which evidently depend on it for an appropriate
motility (116).
PMCA 4 was also found to be crucial in modulating calcium signals in B lymphocytes.
Triggering of the B cell receptor (BCR) in null mutant mice caused an enhanced
transient increase in cytosolic Ca2+, which was attenuated by CD22 (a B cell restricted
transmembrane glycoprotein) via stimulation of Ca2+efflux. The observation suggests a
direct interaction between PMCA 4, CD22 and SHP-1 tyrosine phosphatase (117).
The analysis of the phenotypes of PMCA 2 knockout mice has revealed interesting
characteristics. Though the animals appeared fairly normal at birth, they started to
- 41 -
present balance impairment around day 10 (118) Recording of the auditory brain
response revealed that they were deaf and the analysis of the vestibular inner ear
showed the absence of otoconia. It was also observed that sensory hair cells started to
degenerate after day 10. The most severely affected animals also presented partial loss
of nerve cells. Unexpectedly, Purkinje neurons, even if they abundantly express PMCA
2, were unaltered, and the cerebella maintained normal organization and cell
distribution. Thus, the balance defect in knockout mice was not due to a general
(cerebellar) neuronal impairment but it was rather determined by the absence of
otoconia that sense gravity and equilibrium changes in the vestibular system.
Similar phenotypes were observed in mice presenting naturally occurring mutations of
the PMCA 2 gene. The first historically identified animals with mutated sequence of the
pump were the deafwaddler mice (88). The deafwaddler mutation will be described in
more detail in the next section, but is mentioned here because one of the deafwaddler
mutations generates a null mutant. Indeed, deafwaddlers carry either a point mutation
(G283S) in the first intracellular loop of the PMCA 2 protein (dfwj) or a frame shift
mutation at codon 471 generating a null mutant (dfw2j, dfw3j). Another point mutation
of the PMCA 2 gene (K412E) was identified by a later study in the deaf Wriggle
Sagami mouse. The aminoacidic substitution in the fourth transmembrane domain
determined an altered membrane targeting of the pumps that was therefore no longer
expressed in the stereocilia (119).
The PMCA 2 knockout mouse (and the PMCA 2 mutations) thus shows that this
isoform has a crucial role in the auditory system. Interestingly, in the outer hair cells
PMCA 2 is specifically and selectively expressed in the stereocilia. PMCA 1 and
PMCA 4 are also present in these sensory cells, but are confined to the basolateral
section of the plasma membrane.
4.7.2- HEREDITARY DEAFNESS AND OTHER DISEASE CONDITIONS.
Studies aimed at characterizing disease features caused by genetic defects of the PMCA
proteins, as well as by variations of gene transcription levels have now begun to appear
(120), but for a long time no human disease was directly linked to a defect in any of the
genes of the four different isoforms of the PMCA pump. Only very recently two
mutations in the gene encoding for PMCA 2 have been identified in humans with
hearing impairment. Schultz et al. have described a hypofunctional variant of ATP2B2
- 42 -
that introduced a V586M substitution in the vicinity of the active site of the protein and
enhanced the auditory defect in patients with an accompanying homozygous mutation in
the gene of Cadherin 23 (121). Another mutation in the PMCA 2 gene, leading to a
G293S substitution, has been observed by our group in a child with hearing impairment
in an Italian family (122). It is of interest that the G293S mutation involves the same
mutation only 10 residues downstream the site of the deafwaddler mutation. These
findings, together with the results of the studies on knockout mice for the ATP2b2 gene
described in the preceding section, highlight the essential role of PMCA2 in mammalian
hearing. Apparently, no other PMCA isoform can adequately compensate for the loss of
PMCA 2, or of PMCA 2 activity, in sensory hair cells emphatizing the special
properties of PMCA2, and of its splice variant present in the stereocilia (see below).
The PMCA 2 gene mutations identified so far as causative of hereditary deafness are
summarized in Fig. 10. A particularly interesting mutation has recently been found in a
mouse with hereditary hearing in loss. The phenotype of the mouse and the
biochemistry of the mutated pump are presently being analyzed and the results obtained
so far will be described below. The special interest of the mutation lies in the fact that it
affects a transmembrane domain of the pump: Transmembrane domain substitutions are
normally assumed to alter the membrane targeting of the PMCA pumps, yet, the
mutated pump is correctly delivered to the plasma membrane.
In most cell types non-genetic alterations of PMCA activity contribute to a variety of
pathological states. In turn, modifications of expression levels or abnormal proteolysis
of PMCA have been observed in different organs and systems. In neuronal
differentiation, for instance, up-regulation of calcium channels to trigger calcium
dependent events is associated with PMCA up regulation, to respond to the increased
demand for calcium extrusion (123). In particular, PMCA 2 and 3 isoforms, specifically
expressed in the central nervous system, are up regulated during the development of
granule neurons of cerebellum (124). Isoforms PMCA 2 and PMCA 3 are normally up-
regulated in neuronal development, as shown in the neuroblastoma cell line IMR-32
(125). PMCA 4 is less involved in neuronal differentiation, and may even be down-
regulated (126), whereas PMCA1 is a determinant for neurite extension as shown in
PC-6 neuronal cells (127), where knocking down of the PMCA 1 gene by siRNA leads
to the loss of nerve growth factor-mediated neurite elongation (128).
- 43 -
Figure 10: Deafness-associated mutations of Plasma Membrane Ca2+ ATPase
Alterations of calcium homeostasis play a particularly important role in brain aging,
brain damage by ischemia/anoxia, and several neurodegenerative diseases. Free radicals
affect membrane Ca2+ transport by acting directly on the PMCA or by deranging the
lipidic bilayer that hosts it. For instance, in Alzheimer patients’ neurons, calcium
homeostasis is altered. The amyloid beta-peptide (Abeta), which accumulates in AD
brain, is neurotoxic: its addition to cultured hippocampal neurons caused altered
calcium homeostasis preceded by the impairment of Na-K ATPase and Ca ATPase
activities (129).
Cardiac hypertrophy and myopathy may also be associated with abnormalities in
calcium homeostasis linked to modifications of the PMCA pump. Sarcolemmal
membranes from rats with pressure and hypertension-related cardiac hypertrophy
presented increased activity of the PMCA. This is likely to be an adaptive reaction to
facilitate the removal of Ca2+ from myocardial cells during the development of
hyperthropy (130). In atherosclerosis and restenosis the regulation of apotosis is
controlled by the transcription factor C-myb, whose binding to the PMCA 1 gene
promoter represses pump transcription in vascular smooth cells muscle at G1/S cell
cycle interface (131).
- 44 -
Pancreatic islet and beta cells of pancreas contain a unique combination of PMCA
isoforms, predominantly PMCA 1b, 2b, 3a, 3c and 4a. Evidently, functionally distinct
PMCA isoforms participate in Ca2+ homeostasis in insulin secreting cells (132,133).
The PMCA activity of pancreatic islets is suppressed by high glucose, and is therefore
altered in diabetes (134).
In kidney, the expression of PMCA isoforms can be selectively modulated by
pathophysiological stimuli. At variance with the constant expression level of PMCA1
and PMCA4 isoforms, the expression of PMCA 2 and PMCA 3 in tubules has been
shown to be depressed in hypercalciuric rats (inducing by feeding with a low phosphate
diet) in comparison with hypocalciuric rats. In brain and liver, however, PMCA 2 and
PMCA3 were unchanged, suggesting a specific role for these isoforms in tubular Ca2+
reabsortion (135, 136). Familial hypercalcemia has been associated with global defects
of PMCA function (137).
Finally, PMCA 2 mRNA has been detected in higher level in some breast cancer cell
lines, where it may be overexpressed by a factor of up to 100 fold.
- 45 -
Experimental procedures
1- Baculovirus Expression System
Baculoviruses are the most prominent viruses known to affect the insect population.
They are double-stranded, circular, supercoiled DNA molecules in a rod-shaped capsid.
(Summers, M. and Anderson, D. (1972) J. Virol. 9, 710.) Two of the most common
isolates used in foreign gene expression are Autographa californica multiple nuclear
polyhedrosis virus (AcMNPV) and Bombyx mori (silkworm) nuclear polyhedrosis virus
(BmNPV). Wild-type baculoviruses exhibit both lytic and occluded life cycles that
develop independently throughout the three phases of virus replication. The following
are characteristics of the three phases:
1. Early Phase: In this phase (also known as the virus synthesis phase), the virus
prepares the infected cell for viral DNA replication. Steps include attachment,
penetration, uncoating, early viral gene expression, and shut off of host gene expression.
Actual initial viral synthesis occurs 0.5 to 6 h after infection.
2. Late Phase: In this phase (also known as the viral structural phase), late genes that
code for replication of viral DNA and assembly of virus are expressed. Between 6 and
12 h after infection, the cell starts to produce extracellular virus (EV), also called non-
occluded virus (NOV) or budded virus (BV). The EV contains the plasma membrane
envelope and glycoprotein (gp) 64 necessary
for virus entry by endocytosis. Peak release of extracellular virus occurs 18 to 36 h after
infection.
3. Very Late Phase: In this phase (also known as the viral occlusion protein phase), the
polyhedrin and p10 genes are expressed, occluded virus (OV)—also called occlusion
bodies (OB) or polyhedral inclusion bodies (PIBs)—are formed, and cell lysis begins.
Between 24 and 96 h after infection, the cell starts to produce OV, which contains
nuclear membrane envelopes and the viral polypeptides gp41 and gp74. Multiple
virions are produced and surrounded by a crystalline polyhedra matrix. The virus
particles produced in the nucleus are embedded within the polyhedrin gene product and
a carbohydrate-rich calyx.
- 46 -
1.1- Baculoviruses as Expression Vectors
The major difference between the naturally occurring in vivo infection and the
recombinant in vitro infection is that the naturally occurring polyhedrin gene within the
wild-type baculovirus genome is replaced with a recombinant gene or cDNA. These
genes are commonly under the control of polyhedrin and p10 promoters. In the late
phase of infection, the virions are assembled and budded recombinant virions are
released. However, during the very late phase of infection, the inserted heterologous
genes are placed under the transcriptional control of the strong AcNPV polyhedron
promoter. Thus, recombinant product is expressed in place of the naturally occurring
polyhedrin protein. Usually, the recombinant proteins are processed, modified, and
targeted to the appropriate cellular locations.
1.2- Cytopathogenesis
As the recombinant infection advances, several morphological changes take place
within the cells. The timing of the infection cycle and the changes in cell morphology
vary with the insect cell line and strain of baculovirus used. The metabolic condition of
the culture and growth medium used also can affect the timing of baculovirus infection.
The following morphological changes are typical of monolayer Spodoptera frugiperda
(Sf9) cells infected with recombinant AcNPV.
1. Early Phase: Infection begins with the adsorptive endocytosis of one or more
competent virions by a cell in a high metabolic state (peak replication rate). The
nucleocapsids pass through the cytoplasm to the nucleus. When the virions enter the
nucleus, they release the contents of the capsid. Within 30 min of infection, viral RNA
is detectable. Within the first 6 h of infection, the cellular structure changes, normal
cellular functions decline precipitously, and early-phase proteins become evident.
2. Late Phase: Within 6 to 24 h after infection, an infected cell ceases many normal
functions, stops dividing, and is logarithmically increasing production of viral genome
and budded virus. The virogenic stroma (an electron dense nuclear structure) becomes
well developed. Infected cells increase in diameter and have enlarged nuclei. The cells
may demonstrate reduced refractivity
under phase contrast microscopy. Infected cultures stop growing.
3. Very Late Phase: Within 20 to 36 h after infection, cells cease production of budded
virus and begin the assem-bly, production, and expression of recombinant gene product.
- 47 -
In monolayer cultures, areas of infection display decreased density as cells die and lyse.
Likewise, in suspension cultures, cell densities begin to decrease. Infected cells continue
to be increased in diameter and have enlarged nuclei. The cytoplasm may contain
vacuoles, and the nuclei may demonstrate granularity.
1.2- Generating a Recombinant Virus by Homologous Recombination
Using homologous recombination to generate a recombinant baculovirus is outlined in
Fig. 2a. The most common baculovirus used for gene expression is AcMNPV.
AcMNPV has a large (130-kb), circular, double-stranded DNA genome. The gene of
interest is cloned into a transfer vector containing a baculovirus promoter flanked by
baculovirus DNA derived from a nonessential locus—in this case, the polyhedrin gene.
The gene of interest is inserted into the genome of the parent virus (such as AcMNPV)
by homologous recombination after transfection into insect cells.
Typically, 0.1% to 1% of the resulting progeny are recombinant. The recombinants are
identified by altered plaque morphology. For a vector with the polyhedrin promoter, as
in this example, the cells in which the nuclei do not contain occluded virus contain
recombinant DNA. Detection of the desired occlusion-minus plaque phenotype against
the background of greater than 99% wild-type parental viruses is difficult. A higher
percentage of recombinant progeny virus (nearly 30% higher) results when the parent
virus is linearized at one or more unique sites located near the target site for insertion of
the foreign gene into the baculovirus genome.( Kitts, P.A., Ayres, M.D. and Possee,
R.D. (1990) Nucleic Acids Res. 18, 5667., Hartig, P.C. and Cardon, M.C. (1992) J.
Virol. Methods 38, 61) To obtain an even higher proportion of recombinants (80% or
more), linearized viral DNA that is missing an essential portion of the baculovirus
genome downstream from the polyhedrin gene can be used.( Kitts, P.A. and Possee,
R.D. (1993) Biotechniques 14, 810.) These approaches can take more than a month to
purify plaques, amplify the virus, and confirm the desired recombinants.
- 48 -
Figure 2a: Generating a recombinant baculovirus by homologous recombination
1.4-Generating a Recombinant Virus by Site- Specific Transposition
A faster approach for generating a recombinant baculovirus (Luckow, V.A., Lee, S.C.,
Barry, G.F., and Olins, P.O. (1993) J. Virol. 67, 4566.) uses site-specific transposition
with Tn7 to insert foreign genes into bacmid DNA propagated in E. coli. The gene of
interest is cloned into a pFASTBAC vector, and the recombinant plasmid is transformed
into DH10BAC competent cells which contain the bacmid with a mini-attTn7 target site
and the helper plasmid. The mini-Tn7 element on the pFASTBAC plasmid can
transpose to the mini-attTn7 target site on the bacmid in the presence of transposition
proteins provided by the helper plasmid. Colonies containing recombinant bacmids are
identified by antibiotic selection and blue/white screening, since the transposition
results in disruption of the lacZα gene. High molecular weight mini-prep DNA is
prepared from selected E. coli clones containing the recombinant bacmid, and this DNA
is then used to transfect insect cells. The steps to generate a recombinant baculovirus by
site-specific transposition using the BAC-TO-BAC Baculovirus Expression System are
outlined in Fig. 2b.
- 49 -
Figure 2b: Generation of recombinant baculoviruses and gene expression with the BAC-TO-BAC
expression system
2- Cloning and virus generation of PMCA2 z/b and w/a variants
2.1- Construction of the recombinant baculoviruses
The HindIII fragment from pcDNA3-PMCA2 zb and the Not/KpnI fragment from
pmRFP-PMCA2 wa were inserted in pFASTBACI, non-fusion vector (no fusion tags
are present in the vector) and pFASTBAC HtB fusion vector (fusion with the N-
terminal 6xHis tag and the TEV protease cleavage site). Recombinant baculoviruses
were produced in Sf9 insect cells by the BAC-TO-BAC™ system (Gibco-BRL, Life
Technology Inc., Basel, Switzerland).
2.2- Plasmid DNA isolation from bacteria. Miniprep
E. Coli cells carrying the plasmid of interest were incubated overnight at 37°C at
constant shaking (200-220 rpm) in 5 ml of Luria-Bertani (LB) broth (10g/L Bacto
Tryptone, 5g/L Bacto yeast extract, 10g/L NaCl) supplemented with the appropriate
antibiotic (ampicillin 75 µg/µL or kanamycin 40 µg/µL). The cells were harvested by
centrifugation at 13000 rpm (microcentrifuge Biofuge, Haeraeus) for 1 min and the
plasmid DNA was isolated using the “NucleoSpin Plasmid” protocol (Macherey-Nagel).
- 50 -
The protocol is based on SDS/alkaline lysis of the E. Coli cells followed by
neutralization of the resulting lysate, binding of plasmid DNA to silica membranes
columns and elution under low ionic strength conditions of pure plasmid DNA.
2.3- Plasmid DNA isolation from bacteria. Maxiprep
E. Coli cells carrying the plasmid of interest were incubated overnight at 37°C in 300
ml of LB medium supplemented with antibiotics as described previously. The cells were
harvested by centrifugation at 5000 rpm in a Beckman Coulter centrifuge (rotor JA10)
for 15 min at 4°C. Plasmid DNA was extracted using Large Scale Plasmid MaxiPrep
(Qiagen). The procedure employs an SDS/alkaline method to prepare cleared lysate,
followed by neutralization of the lysate and plasmid DNA binding to anion exchange
resins. The purified plasmid DNA was eluted from the column and concentrated by an
alcohol precipitation. The concentration and quality of the purified DNA was measured
with a UV spectrophotometer at OD260-280
2.4-DNA digestion and agarose gel electrophoresis
Digestion of DNA was performed with restriction endonucleases which recognize unique
DNA palindromic sequences at which they hydrolyze the phosphoester linkage.
An analytical restriction enzyme reaction contained:
- 1 µg DNA (in TE 1X buffer: 10 mM Tris-HCl, 1 mM EDTA, pH 8)
- 2 µl 10x buffer (depending on the restriction enzyme)
- 1 µl restriction enzyme (10 units/µl activity)
- Sterile H2O up to a final volume of 20 µl
The reaction mix was incubated for 1 hour on a water bath at 37°C.
A preparative restriction enzyme reaction contained:
- 3-10 µg DNA (in TE 1X)
- 10 µl 10X buffer mix
- 3 µl restriction enzyme (10 units/µl activity)
- Sterile H2O up to a final volume of 100 µl
The reaction mix was incubated for 2 hours on a water bath at 37°C.
The DNA samples were loaded on 1% agarose (Type I Low EEO, Sigma) gels together
with 1x loading buffer (0.04% bromophenol blue, 0.04% xylene cyanol, 5% glycerol in
H2O) and ran at 80 V for approximately 30 min.
- 51 -
The size of the separated DNA strands was determined by comparison of their relative
position to that of the DNA strands of the 1000bp DNA Ladder (MBI Fermentas).
2.5- DNA fragments purification from agarose gels
Following agarose gel electrophoresis, the gel slice containing the desired DNA band
was excised with a sharp scalpel and purified using the “NucleoSpin Extract” kit
(Macherey-Nagel). The protocol is based on DNA binding to a silica membrane in the
presence of chaotropic salts and elution of pure DNA under low ionic strength conditions
with TE 1x buffer. Contaminations like salts and soluble macromolecular components
are prevented by a washing step with an ethanolic buffer.
2.6- DNA ligation
DNA ligation was performed by incubating the DNA fragments with the appropriately
linearized cloning vector in the presence of 2-4 units of T4 DNA ligase (3 units/µL), 10x
T4 ligase buffer (50 mM TrisHCl pH 7.5, 10 mM MgCl2, 10 mM DTT, 25 µg/ml BSA, 1
mM ATP) and sterile water to a final volume of 10 µL . A vector: insert molar ratio of
1:3 was usually used.
The reaction mix was incubated overnight at 16°C
2.7- Transformation of competent E. Coli cells
TOP10 chemically competent cells (Invitrogen) which have been kept on freezer storage
were thawed on ice.1-200 ng of plasmid DNA were added to the competent cells and the
transformation mix was kept on ice for 30 min. The cells were heat-shocked for 30-40
seconds at 42° and then incubated for 30 min at 37°C in 500 µl of LB broth. The mix
was plated out on LB agar which contained the antibiotic ampicillin that selects for
transformants.
2.8- DNA sequencing
DNA sequencing was performed by the BMR Genomics Center of the University of
Padova.
- 52 -
2.9- Generation of the Recombinant Bacmid
The baculovirus shuttle vector (bacmid), bMON14272 (136 kb), present in DH10Bac E.
coli (Invitrogen) contains:
• A low-copy number mini-F replicon
• Kanamycin resistance marker
• A segment of DNA encoding the LacZα peptide from a pUC-based cloning vector into
which the attachment site for the bacterial transposon, Tn7 (miniattTn7) has been
inserted. Insertion of the mini-attTn7 does not disrupt the reading frame of the LacZα
peptide.
The bacmid propagates in E. coli DH10Bac as a large plasmid that confers resistance to
kanamycin and can complement a lacZ deletion present on the chromosome to form
colonies that are blue (Lac+) in the presence of a chromogenic substrate such as Bluo-
gal or X-gal and the inducer, IPTG.
Recombinant bacmids (composite bacmids) are generated by transposing a mini- Tn7
element from a pFastBac donor plasmid to the mini-attTn7 attachment site on the
bacmid. The Tn7 transposition functions are provided by a helper plasmid
2.10- Transformation of the DH10Bac E. coli cells
For transformation of MAX Efficiency DH10Bac chemically competent cells with the
pFastBac construct, positive controls for transposition were included (i.e. pFastBac
expression plasmid).
1. Thaw on ice one vial of MAX Efficiency® DH10Bac™ competent cells for each
transformation.
2. For each transformation, 100 µl of the DH10Bac cells were gently mixed and
transferred into a pre-chilled, 15 ml round-bottom polypropylene tube.
3. Appropriate amount of plasmid DNA was added into the DH10Bac cells.
• pFastBac construct: 1 ng (5 µl)
• pFastBac™ control plasmid: 1 ng
• pUC19 control: 50 pg (5 µl)
4. The cells were incubated on ice for 30 minutes and were subjected to heat shock for
45 seconds at 42°C without shaking
5. The tubes were transferred into the ice and chilled for 2 minutes
- 53 -
6. 900 µl of S.O.C. Medium (Bacto-tryptone, 20 g/L, Bacto-yeast extract, 5 g/L, NaCl,
0.5g/L, 1M KCl, 2.5 ml/L) was added at room temperature.
7. For pFastBac™ transformations: tubes were shacked at 37°C at 225 rpm for 4
hours.
For pUC19 transformation: at 37°C at 225 rpm for 1 hour.
For each pFastBac™ transformation: 10-fold serial dilutions of the cells were
prepared (10-1,
10-2, 10-3) with S.O.C. Medium. Plated 100 µl of each dilution on an LB agar plate
containing 50 µg/ml kanamycin, 7 µg/ml gentamicin, 10 µg/ml tetracycline, 100 µg/ml
Bluo-gal, and 40 µg/ml IPTG.
For the pUC19 transformation: the cells were diluted 1:100 with S. O.C. Medium.
Plated 100 µl of the dilution on an LB agar plate containing 100 µg/ml ampicillin.
8. Plates were incubated for 48 hours at 37°C. White colonies were picked for analysis
(Insertions of the mini-Tn7 into the mini-attTn7 attachment site on the bacmid disrupt
the expression of the LacZα peptide, so colonies containing the recombinant bacmid are
white in a background of blue colonies that harbor the unaltered bacmid.) (Fig. 2c)
9. White colonies were picked and restreaked on fresh LB agar plates containing 50
µg/ml kanamycin, 7 µg/ml gentamicin, 10 µg/ml tetracycline, 100 µg/ml Bluo-gal, and
40 µg/ml IPTG. The plates were incubated overnight at 37°C. From a single colony was
confirmed to have a white phenotype on restreaked plates containing Bluo-gal and
IPTG, were inoculated the liquid culture containing 50 µg/ml kanamycin, 7 µg/ml
gentamicin, and 10 µg/ml tetracycline.
10. Recombinant bacmid DNA was isolated using the Marligen MiniPrep Kit (Catalog
no. 11449-022)
- 54 -
Figure 2c: Recombinant Bacmids colonies selected with X-gal: A. Different diluted dishes with recombinant bacmid colonies, the negative colonies are blue; B. A zoom in of a dish where is possible distinguish blue colonies from white recombinant colonies
2.11- Analysis of Recombinat bacmid DNA by PCR
The bacmid contains M13 Forward (-40) and M13 Reverse priming sites flanking the
mini-attTn7 site within the lacZα-complementation region to facilitate PCR analysis.
(Fig. 2d)
Primers sequences:
M13 Forward (-40) 5′d [GTTTTCCCAGTCACGAC]3′
M13 Reverse 5′d [CAGGAAACAGCTATGAC]3′
Figure 2d: PCR analysis of Recombinant bacmid DNA
For each sample, was set up the following 50 µl PCR reaction in a 0.5 ml
microcentrifuge tube:
A BA B
- 55 -
-Recombinant bacmid DNA (100 ng) 1 µl
-10X PCR Buffer (appropriate for enzyme) 5 µl
-10 mM dNTP Mix 1 µl
-50 mM MgCl2 1.5 µl
-PCR Primers (1.25 µl each 10 µM stock) 2.5 µl
-Sterile Water 38.5 µl
-Platinum Taq polymerase (5 units/µl) 0.5 µl
-Total Volume 50 µl
The PCR was amplified using the following cycling parameters:
Initial Denaturation : 3 minutes 93°C 1X
Denaturation: 45 seconds 94°C
Annealing: 45 seconds 55°C 25-35X
Extension: 5 minutes 72°C
Final Extension: 7 minutes 72°C 1X
5-10 µl from the reaction was removed and analyzed by agarose gel electrophoresis.
2.12- Transfection of Insect cells
The Bacmid transfection was performed using Cellfectin reagent (Invitrogen). This
reagent is a 1:1.5 (M/M) liposome formulation of the cationic lipid N,NI, NII, NIII-
Tetramethyl-N, NI, NII, NIII-tetrapalmitylspermine (TM-TPS) and dioleoyl
phosphatidylethanolamine (DOPE) in membrane-filtered water.
The transfection was assayed as followed:
1. 9 x 105 per well insect cells was seeded in a 6-well or 35 mm tissue culture plate with
2 ml of growth medium containing antibiotics (e.g. 2 ml of Sf-900 II SFM medium
(GIBCO) containing 1 µg/ml gentamicin final concentration).
2. The cells were allowed to attach at 27°C for at least 1 hour.
3. For each transfection sample was prepared bacmid DNA:Cellfectin Reagent
complexes as follows in 12 x 75 mm sterile tubes.
a. 1 µg of purified bacmid DNA was diluted in 100 µl of unsupplemented Grace’s
Medium (GIBCO).
b. Cellfectin Reagent was mixed thoroughly before use by inverting the tube 5-10
times. 6 µl of reagent was removed and diluted in 100 µl of unsupplemented Grace’s
Medium.
- 56 -
c. The diluted bacmid DNA with the diluted Cellfectin Reagent were combined (total
volume was ~210 µl) and incubated for 15 to 45 minutes at room temperature.
4. While DNA: lipid complexes were incubated, the media from the cells was removed
and washed once with 2 ml of unsupplemented Grace’s Medium.
5. 0.8 ml of unsupplemented Grace’s Medium was added to each tube containing the
DNA:lipid complexes DNA:lipid complexes were mixed gently and added the to each
well containing cells.
6. The cells were incubated in a 27°C incubator for 5 hours.
7. The DNA:lipid complexes were removed and added 2 ml of complete growth media
(Sf-900 II SFM containing antibiotics) to the cells.
8. The cells were incubated in a 27°C humidified incubator for 5 days or 7 days.
. At 5th day post transfection, Vo virus was produced and isolated from medium.
2.13- Virus amplification
A volume of 100 ml of Sf9 insect cells, with a density of 2X106cells/ml, were infected
with the viral particles, using a multiplicity of infection (MOI) of 0.1 to1. MOI is
defined as the number of virus particles per cell. The MOI was calculated using the
following formula:
Inoculum required (ml) = MOI (pfu/cell) x number of cells
titer of viral stock (pfu/ml)
After 48 hour incubation, cells were sedimented at 1300 rpm for 10 minutes at 4°C. The
supernatant, containing the virus, was filtered to eliminate Sf9 cell residues and stored
at 4°C.
3- Expression and purification of PMCA2 zb and wa variants
3.1 - Infection and expression of PMCA2 in insect cells (Figure 2e)
To express PMCA 2 isoforms a suspension cell culture of Sf9 insect cell line was used.
The cell growth was carried out at 27°C in Sf-900 II SFM medium with L-Glutamine
(GIBCO), 10% (v/v) Foetal Bovine Serum (GIBCO) 1 µg/ml Gentamicin (GIBCO).
- 57 -
When the cells in the medium reached the mid-logarithmic phase of growth,
corresponding to a density of 1X106 to 2X106cells/ml, cells were ready to be infected
with the virus.
Infection was performed using a multiplicity of infection (MOI) of 1 to 5.
A time course was performed to determine the expression kinetics for the protein,
considering also the incidence of the degradation process as long as the expression
proceeds.
The expression was finally carried out at 27°C by infecting 1 litre of culture. Cells were
harvested 18 hour after the infection by centrifugation at 1500 rpm with a JA14 Rotor in
an Avanti J-20 XP Centrifuge (Beckman Coulter) for 10 min at 4°C.
Figure 2e : Morphological differences between uninfected and infected sf9 cells
A. Uninfected Sf9 cells continue to divide and form a confluent monolayer B. Sf9 cells infected with recombinant baculovirus, these cells stop dividing and enlarge
3.2- Sf9 cell membrane preparation
Sf9 cells from a litre of culture were washed three times with 25 mM Tris pH 7.5, 0.15
M NaCl, 0.1 mM PMSF, centrifuging each time at 1500 rpm with a JA14 Rotor (Avanti
J-20 XP Centrifuge, Beckman Coulter) for 10 min at 4°C.
The pellet was resuspended in 20 ml of 10 mM Tris pH 7.5, 1 mM DTT and Protease
Inhibitors (Cocktail Tablets complete EDTA-free (Roche).
The resuspended pellet was then homogenised with an Ultra Turrax T25 Basic (IKA
Labortechnic) to lyse the cells. To the homogenised material Sucrose to 10% (w/v) and
KCl 150 mM final concentration were added.
A
B
- 58 -
The lysed cells were centrifuged at 750g for 10 min at 4°C with a centrifuge Z383K
(Hemle). To the supernatant EDTA to a 10 mM final concentration was added. The
supernatant was finally centrifuged at 120000Xg for 45 min at 4°C.
The pellet was resuspended in 4 mM Tris-HCl pH 7.5, 10% (w/v) Sucrose and Protease
Inhibitors.
3.3- Solubilisation of PMCA2 proteins
Sf9 isolated membranes were treated with an amount of detergent C12E8 equal to the
total protein amount within the membranes.
Membranes were incubated with the detergent at 4°C for 1 hour in rotation.
Solubilised material was then centrifuged at 120000Xg for 30 min at 4°C.
To the supernatant were added 0.5 mg/ml Phosphatidylcholine, 15% (v/v) Glycerol, 3
mM DTT, 50 µM CaCl2.
3.4- Calmodulin affinity chromatography
Membrane proteins extract was incubated at 4°C for 1 hour rotating with a Calmoduline
Sepharose 4B (Amersham Biosciences) opportune bed volume, previously equilibrated
with 20 mM Hepes
pH 7.2, 120 mM KCl, 1 mM MgCl2, 100 µM CaCl2, 0.32% (w/v) C12E8, 0.5 mg/ml
Phosphatidylcholine, 15% (v/v) Glycerol, 3 mM DTT.
The resin with the bound protein is then harvested by centrifugation at 2000Xg in a
centrifuge Z383K (Hemle) for 3 min at 4°C, and the supernatant (flow through) is
conserved.
The resin was washed with the equilibration buffer and the protein was finally eluted by
incubation at 4°C for 30 min rotating in 20 mM Hepes pH 7.2, 120 mM KCl, 1 mM
MgCl2, 0.05% (w/v) C12E8, 0.5 mg/ml Phosphatidylcholine, 15% (v/v) Glycerol, 10 mM
EDTA, 3 mM DTT.
The resin was centrifuged at 2000 Xg for 3 min at 4°C and to the supernatant,
containing PMCA2zb were added 1.5 mM MgCl2 and 4.8 µM CaCl2.
The protein concentration of the samples from the membrane preparation and the
protein purification was determined by using the Lowry method.
- 59 -
3.5- His Tag affinity purification
PMCA 2 isoforms were subjected to His GraviTrap (GE Healthcare). These are a
prepacked, single-use gravity-flow column containing precharged Ni Sepharose 6 Fast
Flow. The column is intended for purification of histidine-tagged proteins by
immobilized metal affinity chromatography (IMAC). Ni Sepharose 6 Fast Flow has
high protein binding capacity, low nickel ion (Ni2+) leakage and is compatible with
denaturing agents plus a wide range of additives. Both clarified and unclarified sample
can be applied to the column, and special frits protect the medium from running dry
during purification
Solubilised membrane proteins were applied to the prepacked columns at 4°C for 2,
previously equilibrated with 20 mM Hepes pH 7.2, 300 mM KCl, 500 µM MgCl2, 100
µM CaCl2, 3 mg/ml C12E8, 0.5 mg/ml Phosphatidylcholine, 15% (v/v) Glycerol, 1 mM
DTT.
The columns with the bound protein is then harvested by centrifugation at 1000 rpm in a
centrifuge Z383K (Hemle) for 3 min at 4°C, and the flow through was conserved.
The His GraviTrap was washed with the washing buffer (20 mM Hepes pH 7.2, 300
mM KCl, 500 µM MgCl2, 100 µM CaCl2, 0.5 mg/ml C12E8, 0.5 mg/ml
Phosphatidylcholine, 15% (v/v) Glycerol, 1 mM DTT and 10 mM Imidazole) and the
protein was finally eluted by incubation at 4°C for 30 min rotating in the same washing
buffer but with different concentration of Imidazole (25, 50 and 100 mM ).
At the eluted protein was added 1.5 mM MgCl2 and 4.8 µM CaCl2.
The protein concentration of the samples from the membrane preparation and the
protein purification was determined by using the Lowry method.
3.6- PMCA2 preparation for SDS-PAGE analysis
Proteins were separated on SDS-PAGE essentially in accordance with the Laemmli
method (Laemmli, 1970). Separating gels of 7.5 or 10 %, depending on protein size, and
stacking gels of 5% were used for the electrophoretic separation of proteins. The running
buffer contained 25 mM Tris-HCl pH 8.3, 192mM Glycine and 0.1% SDS and the
electrophoresis equipment was from Amersham Bioscience. The gels were run at 25 mA
for approximately 2-3 hours.
- 60 -
SDS-PAGE was also conducted on NuPAGE®NOVEX pre-cast polyacrylamide mini-
gels from Invitrogen. The NuPAGE® System is based upon a Bis-Tris-HCl [Bis (2-
hydroxyethyl) imino-tris (hydroxymethyl) methane-HCl)] buffered (pH 6.4)
polyacrylamide gel, with a separating gel that operates at pH 7.0. The NuPAGE MOPS
SDS running buffer (20X) contained 50 mM MOPS [[3-(N-morpholino)] propane
sulfonic acid], 50 mM Tris base, 0.1% SDS and 1mM EDTA at pH 7. The gels were run
at 200 V for approximately 50 min.
SeeBlue Plus2 Pre-Stained Standard (Invitrogen), a mix of 34 proteins with molecular
weights from 4 to 250 kDa, was employed for visualizing protein molecular weight
range and evaluating transfer efficiency.
Following electrophoresis, the gel was stained with Coomassie Brilliant Blue, allowing
visualisation of the separated proteins, or further processed for Western blotting analysis.
3.7- PMCA2 Western Blotting
To detect eventual degradation of PMCA2 during the preparation, all the fractions were
analysed by Western Blot, using a PMCA anti-mouse antibody 5F10 (ABR),
Following electrophoresis, the separated proteins were transferred onto PVDF
(polyvinyldene difluoride) membranes (Amersham Bioscience) at 350 mA for 2 h in a
Blotting buffer containing 47.9 mM TrisHCl pH 8.4, 38.6 mM glycine, 0.0385% SDS
and 10% methanol. Blocking of non-specific binding was achieved by placing the
membranes for 1h in a saturating buffer (5% non-fat dry milk and 0.1% Tween 20
detergent in PBS). After blocking, a dilute solution of the primary antibody and the
membranes were incubated together under gentle agitation for 60-90 min at room
temperature, or overnight at 4°C. The membranes were then rinsed for 1h with fresh
saturating buffer in order to remove the unbound primary antibody and were incubated
for 30-45min with the HRP (horseradish peroxidase)-conjugated secondary antibodies
(1:2000 anti-mouse, anti-rabbit or anti-goat IgG HRP, Santa Cruz Biotechnology). The
membranes were rinsed with saturating buffer and then with PBS and the
chemiluminiscence was detected with the ECL techniques (Amersham Biosciences).
The signal was captured on Kodak BioMax light films (Sigma).
- 61 -
3.8- Western blotting with anti Calmodulin antibody
The western blotting with anti-Calmodulin (Sigma Aldrich) antibody was performed
according Sigma’s protocol. The PVDF membranes were incubated for 60 minutes in
Fixation Buffer (2% (v/v) glutaraldehyde, freshly prepared in KP Buffer). The
membranes were rinsed three times in KP buffer (25 mM KH2PO4/K2HPO4 buffer, pH
7.0), then incubated in 1 M Lysine for 60 minutes. Blocking of non-specific binding
was achieved by placing the membranes overnight in a saturating buffer (5% (w/v) BSA
in PBS ). After blocking, a dilute solution of the primary antibody (1:1000 in PBS-
BSA) and the membranes were incubated together for 2 hours or overnight at 4 °C. The
membranes were then rinsed 3 times for 5 minutes each at room temperature using
washing buffer (PBS + 0.05% (v/v) TWEEN 20) and were incubated for 1 hour with the
HRP (horseradish peroxidase)-conjugated mouse secondary antibody (1:2000).
4- Expression of PMCA pumps in CHO cells
4.1- Continuous cell lines
CHO (Chinese Hamster Ovary) derive from epithelial cell chinese hamster ovary cells
and are cultured in 75cm2 Falcon flasks in Ham’s F12 medium (Gibco) supplemented
with 10% fetal bovine serum (FBS), 2mM glutamine, penicillin (60µg/µl) and
streptomycin (120µg/µl)
The cells were maintained in a Haereus incubator at 37°C, in a humid atmosphere
containing 5% CO2 and were split every 2-3 days. When the cells reached 70-80% they
were detached from the flask using 0,05% trypsine/0,2%EDTA, in PBS (140mM NaCl,
2mM KCl, 1,5mM KHPO4, 8mM Na2HPO4, pH=7,4). Fresh medium was added to the
flask in order to inhibit the trypsinization and the cells were collected by centrifugation
at 1000 rpm for 3 min, 25°C in a Beckman GS-15R centrifuge (rotor S4180). The pellet
was resuspended in 5ml of medium and the cells were seeded onto 13mm glass cover
slips or on 24-multiwell plates (for Western blotting analysis) and allowed to grow to
50% confluence. The cells were used for transfection experiments, Western Blot
analysis and preparation of membranes cells.
- 62 -
4.2- Transfection
CHO cell lines were transfected using the calcium-phosphate method. One day prior to
transfection, the cells were plated onto 150x25 mm dishes such that they were
logarithmically growing on the day of transfection (i.e. 30-40% confluent at the time of
transfection). 30 µg of plasmid DNA were used for every single dish. The transfection
mix was prepared as follows: 25µl of CaCl2 2.5 M were added to the plasmid
DNA/TE1x 225µl total mix. The tube containg 250µl of HBS 2x (280mM NaCl, 50mM
HEPES, 1.5mM Na2HPO4, pH 7.12) was gently vortexed and the DNA/TE/CaCl2 mix
was added drop wise. A fine opalescent DNA-calcium-phosphate precipitates formed in
about 10 minutes and the transfection mix was then added to the cells (in fresh DMEM
growth medium that was changed 2 h before transfection). After 14-16h the cells were
washed 2 times with PBS and 1 time with the appropriate growth medium in order to
eliminate the DNA-calcium-phosphate precipitates that did not enter the cells. The cells
were incubated at 37°C until use.
4.3- Isolation of Microsomes from CHO Cells
.
Cells from five 150x25 mm dishes were washed once with 20 ml of phosphate-buffered
saline containing 1 mM EDTA and harvested in 10 ml of phosphate-buffered saline
containing 0.1 mM phenylmethylsulfonyl fluoride and Cocktail proteases inhibitors
EDTA-free (Roche). Cells were collected by centrifugation ( 2,000 X g, 10 min) at 4 0C
and resuspended in 6 ml of a hypotonic solution of 10 mM Tris- HCl, pH 7.5, 1 mM
MgC12, 0.1 mM phenylmethylsulfonyl fluoride, Cocktail proteases inhibitors EDTA-
free, and 2 mM dithiothreitol. The cells were swollen for 15 min on ice and then
subjected to three cycles of freeze and thaw. The homogenate was diluted with an equal
volume of 0.5 M sucrose, 0.3 M KCl, 2 mM dithiothreitol, 10 mM Tris-HC1, pH 7.5,
homogenized again with N2 liquid, and centrifuged at 5,000 X g for 15 min. The
supernatant was made 0.6 M in KC1 and, in order to remove calmodulin, an excess of
EDTA (1.5 mM) was also added. The suspension was centrifuged at 100,000 X g for 40
min to sediment the microsomal fraction. The final pellet was resuspended in a solution
containing 0.25 M sucrose, 0.15 M KC1, 10 mM Tris-HC1, pH 7.5, 2 mM
dithiothreitol, and 20 µM CaCl2, at a protein concentration of 1-3 mg/ml and stored in
liquid N2 until needed.
- 63 -
4.4- Mutagenesis
In vitro site-mutagenesis was carried out with QuikChange II Site-Directed Mutagenesis
Kit (Stratagene) according to the manufacturer's instructions.
5- ATPase activity The ATPase activity was measured by Coupled Enzyme Assay.
The coupled enzyme ATPase assay is based on the conversion of phosphoenolpyruvate
(PEP) to pyruvate by pyruvate kinase (PK) coupled to the conversion of pyruvate to
lactate by lactate dehydrogenase (LDH). The latter step requires NADH which is
oxidized to NAD+. NADH absorbs strongly at 340 nm but NAD+ does not, enabling the
ultilization of NADH to be followed by monitoring absorbance at 340 nm. The decrease
in OD340 can be converted into ATPase activity where 1 molecule of NADH oxidized to
NAD+ corresponds to the production of 1 molecule of ADP by the motor ATPase. (Fig.
2f)
Figure 2f: Coupled Enzyme Assay The assay was carried out at 37°C in a final volume of 1 ml of a mixture containing precise amounts of reagents reported in the table below:
- 64 -
Reagents
Final Concentration
Tris pH 7.2 20 mM MgCl2 5 mM EGTA 0.5 mM KCL 0.1 M PEP 0.5 mM NADH 0.15 mM PK/LDH (Roche) 1.4 U/ml ATP 40 mM PMCA Membranes 25 µg/ml CaCl2 50 mM
Table II. Coupled Enzyme Assay Medium components.
The ATPase activity, detected at 340 nm (DU640 Spectrophotometer, Beckman
Coulter), is expressed in µmol Pi/min*µg of Protein (Pyrophosphate moles originated
by the ATP hydrolysis): the maximal activity and the basal activity are calculated by
multiplication of the activity curve slope value by a factor considering the NADH molar
extinction coefficient (εNADH) and the amount of protein (µg). The real activity was
obtained by subtracting the basal activity from the maximal activity. The assay was
performed in presence of ouabain, thapsigargin and 25 µM PhosphatidylSerine (Sigma
Aldrich) and 200 nM Calmodulin (Sigma Aldrich).
6- Immunocytochemistry
Thirty-six hours after transfection, CHO cells were processed for immunofluorescence
as follows: they were washed twice with PBS, fixed for 15 min in 3.7% formaldehyde,
washed three times with PBS and then incubated for 10 min in PBS supplemented with
50mM NH4Cl. Permeabilization of membranes was obtained by incubating the cells for
5 min with 0.1% Triton X-100 in PBS, followed by a 1 h wash with 1% BSA (Sigma
Aldrich) in PBS. After 2-3 washes with PBS, the cells were incubated for 1h at room
temperature in a wet chamber with a primary monoclonal antibody against PMCA
(ABCAM) or against GFP (Molecular Probes) , at a 1:100 dilution in PBS. Staining was
then carried out with the secondary antibody Alexa Fluor 488 (Molecular Probes) at a
1:200 dilution in PBS. After each incubation, the cells were washed four times with
- 65 -
PBS. The images were acquired using a spinning disk confocal microscope Ultraview
(Perkin Elmer Life Sciences) equipped with a 60X oil immersion objective 1.4
N.A.(PlanApo, Nikon).
7- Computational Methods
7.1- Sequence Alignment
A BLAST search performed using the Protein Data Base identifiy Rabbit
SarcoEndoplasmic Reticulum Calcium ATPase (SERCA) pump 1 as the only protein
homologous to PMCAs whose structure has been determined. Therefore, the primary
sequence of the zb variant of the human PMCA2 (Uniprot entry code: Q01814) was
aligned against that of the SERCA1 pump using T-COFFEE (at www.igs.cnrs-
mrs.fr/Tcoffee/) (REF Notredame, C., Higgins, D. G., and Heringa, J. (2000). T-
Coffee: A novel method for fast and accurate multiple sequence alignment. J.Mol.Biol.
302, 205-217) and then manually refined to take into account secondary structure.
7.2- Structural Model.
Based on this sequence alignment, we constructed a comparative modeling-based 3D
structure of the PMCA. The model was built with Modeller 8.0 with the automatic loop
refinement option (Fiser and Sali 2003a;Fiser and Sali 2003b;Sali and Blundell 1993).
Among the different structures available for the SERCA pump we choose the calcium
bound form (E1 state, REF Toyoshima, C., Nakasako, M., Nomura, H., Ogawa, H.
Crystal structure of the calcium pump of sarcoplasmic reticulum at 2.6 A resolution.
Nature v405 pp. 647-655, 2000) PDB entry code:1SU4) since it presents a more open
and less structured conformation near the gap corresponding to the insertion of the
splicing site A, facilitating the modelling. Residues 35 to 1093 in the sequence of
PMCA2 were used to construct a structural model. The selection of the best model was
based on the Modeller’s objective violation function and WHAT IF analysis (at
http://swift.cmbi.kun.nl/WIWWWI/) (Hooft, R. W., Vriend, G., Sander, C., and Abola,
E. E. (1996). Errors in protein structures. Nature 381, 272). The Ramachandran plot of
the selected model gave 91 % of residues in the most favoured zones, and no residues in
forbidden region suggesting a geometric and energetically stable structure.
- 66 -
RESULTS
I- Expression and purification of PMCA 2 pump.
Functional properties of an unusual isoform of the plasma membrane calcium
ATPase: PMCA 2
Two PMCA pump isoforms (PMCA1 and PMCA4) operate in all tissues and two
(PMCA2 and PCMA3) only in specialized tissues, such as muscle and, especially, brain
(138). All four isoforms display splice variants caused by the insertion of alternative
exons in the primary transcripts: in the translated pump the insertions occur at site A in
the cytosolic loop connecting transmembrane domains 2 and 3 and at site C in the C-
terminal tail of the pump. Alternative splicing is peculiarly complex in PMCA2 as it
involves the insertion of up to three novel exons at site A and of two at site C. The A-
site insertions are in-frame, creating variant w when three exons are inserted (the normal
variant without site- A inserts is termed z). The insert at site C creates instead a novel
stop codon, leading to the truncation of the pump (variants being the normal full-length
pump). The site-C insertions eliminate approximately half of the calmodulin binding
domain; those at site A occur next to a domain that binds activatory acidic
phospholipids, which, however, also bind to the calmodulin binding domain (67). As
expected, the site-C insertions lower the affinity of PMCA pumps for calmodulin (139)
but it is still not clear whether it also compromises the activation by acidic
phospholipids, which is alternative to that by calmodulin (140). It would be plausible to
suggest that the site-A insertion impairs the activation by acidic phospholipids,
particularly in the C-terminally truncated pump variants, in which the C-terminal
phospholipid binding domain also is compromised.
Several splice variants of the four PMCAs have been detected in cochlear cDNAs (141),
but the C-terminally truncated PMCA2a being the only isoform detected in the
stereocilia of hair cells (98). PMCA1b prevails instead in basolateral membranes.
Recent work has shown that the truncated isoform in the stereocilia of the outer hair
cells (OHCs) is also spliced at site A and is thus the wa variant (102). It is plausible to
assume that the choice of this unusual variant of the PMCA (92) was dictated by the
special Ca2+ homeostasis demands of the ambient in which the pump must operate.
We have decided to study the biochemical properties of the PMCA 2 and of its splicing
variant (wa) present in the outer hair cells of the Corti organ of the inner ear.
- 67 -
The first purpose of this work was the PMCA 2 purification in Baculovirus System and
the longer term goal of the project was the preparation of adequate amounts of purified
PMCA 2 to attempt crystallization experiments.
The second step was the characterization of the point mutations in the wa variant pump
that has been shown to induce deafness in mice and humans: thus, they have also been
investigated.
1- Expression of the PMCA2 isoform in the Baculovirus system
1.1-Localization of the PMCA 2 zb in insect cells
It is known that most secretory and plasma membrane proteins require several
processing steps in order to reach the final active form, and that this maturation is
essential for the correct transport of the protein to its final destination. To investigate if
the PMCA2 zb protein was positioned in the plasma membrane, Sf9 cells were infected
with the PMCA2 zb Baculovirus. After 48h cells were fixed in paraformaldehyde and
incubated with the following antibodies: monoclonal antibody against all the
PMCA isoforms (5F10) and polyclonal antibody against the PMCA2 pump. A FITC
conjugated secondary antibody and Alexia 594 were used to produce the fluorescence
(Fig. 3c). In infected cells both 5F10 and 2N recognized the Calcium pump. The
infection score was more or less 80%.
Figure 3c: Inmunofluorescence localization of 5F10 and 2N reactive proteins in Sf9
insect cells A. 5F10 labeled cells infected with PMCA2 zb Baculovirus, showing that the calcium pump is localized preferentially in the plasma membrane B. Image of the 5F10 antibody with 40X objective in which was possible to appreciate the membrane localization. C. Inmunofluorescence membrane localization with 2N antibody.
C
A
B
- 68 -
1.2- Expression and purification of PMCA 2 zb
After that PMCA 2 membrane localization and successfully infection was confirmed;
crude membranes from Sf9 cells infected with PMCA 2 zb baculovirus for 72 hours
were prepared and solubilized; their proteins were separated by SDS-PAGE and stained
with Coomasie Brilliant Blue. In Fig. 3d band of molecular mass of approximately 135-
140 kDa was observed in membranes preparation with/without calcium. The membrane
fractions were richer in PMCA2 protein than total cells.
Figure 3 d: Expression of the PMCA 2 isoforms.The membranes were isolated and separated in 7.5 % SDS-PAGE gel 1. Total sf9 cells 2. Membranes prepared without Ca2+ 3. Sf9 Membranes prepared with calcium. The blue circle shows the PMCA 2 zb expression.
1.3- Time course of PMCA 2 zb expression in Sf9 cells
A PMCA 2 zb expression time course was performed to analyze the time-dependent
dimerization and degradation of PMCA in the Sf9 insect cells. Avoiding dimerization
and degradation is crucial for PMCA binding to the Calmodulin resin for affinity
chromatography purification. Sf9 cells infected with Baculovirus were then collected at
several infection times and the protein concentration was assayed with the Lowry et al.
method (142) (Table III).
kDa 250
148
98
3 1 2
- 69 -
TABLE III: PMCA2 zb expression time course protein determination
The PMCA 2 zb expression was then checked by Western blotting with an antibody
against the PMCA (Fig 3e). PMCA 2 pump appears to be strongly degraded at 40 hours
post infection while at 12 and 18 hours non degradation and dimerization were
observed. Evidently, as the expression time increased Sf9 cells develop proteolytic
systems that degrade the recombinant PMCA 2. An expression time of 18 hours was
thus chosen for the infection prior to the calmodulin affinity purification.
Figure 3e: Time course of the Sf9 insect cells infection to determine the level of the
expression of PMCA 2 z/b. In this figure were indicated the times of infection, the monomer and dimmers of PMCA 2, degradation products were observed after 24 hours post infection.
Infection times Protein concentration
12 hours 6.4 µg/µl
18 hours 6.8 µg/µl
24 hours 7.4 µg/µl
36 hours 7.8 µg/µl
40 hours 8.2 µg/µl
monomer
dimer 250
148
Degradation
products
12 18 24 36 40 hs
kD
- 70 -
1.4- Preparation of Sf 9 cells membranes for purification with a calmodulin resin
The membranes prepared from Sf9 cells infected with recombinant baculovirus were assayed with the Lowry method for protein determination. The protein concentrations of different membranes fractions are represented in the TABLE IV
Preparation membrane
Fraction Protein concentration
Total Cells 20 µg/µl
Supernatant homogenized 10,3 µg/µl
Membranes 9,1 µg/µl
TABLE IV: Sf9 membranes protein determination
The fractions were then checked by Western blotting with an anti PMCA 2 antibody (Fig. 3f). It was confirmed that after 18 hours of infection, the degradation of PMCA 2 zb in the membrane fraction is prevented.
Figure 3f: Immunoblotting of the Sf9 membranes preparation fractions with anti
PMCA antibody 1.Total cells; 2. Supernatant homogenized; 3.Total membranes
1.5- PMCA 2 zb solubilization and purification by a calmodulin affinity resin.
Membranes prepared from Sf9 cells infected with recombinant baculovirus contained a
major band at 135 kDa corresponding to PMCA 2 protein. The membranes containing
the pump isoform were solubilized with a non-ionic detergent C12 E8 at a ratio 1:1. A
calmodulin Sepharose resin was then used. The procedure removed most unrelated
proteins, but contaminants were still visible in Colloidal Coomassie gels of the EDTA
eluates (Fig 3g A).The protein concentrations was determined by the method of Lowry
(TABLE V)
1 2 3
- 71 -
Solubilized Fraction Protein concentration
Membranes solubilized 5 µg/µl
Flow Through 3,3 µg/µl
Elution I 2 mM EDTA 2,75 µg/µl
Elution II 2 mM EDTA 2,8 µg/µl
TABLE V: Membranes solubilization and Calmodulin affinity fractions protein
determination
All fractions were subjected to SDS-PAGE and stained with Colloidal Coomasie, but
proteins in the eluted fractions were only visible after concentration with Vivaspin
concentrator tube. (Fig 3g B)
Figure 3g: Partial purification of the recombinant PMCA 2 by Calmodulin
Affinity
A. 1. Total cells; 2. Supernatant Homog. ; 3. Membranes 4. Pellet solub.; 5- Super. Solub.; 6. FT; 7. elution I 2mM EDTA; 8. elution II 2mM EDTA B. Elution I and II after vivaspin concentration To verify the presence of PMCA 2 in the eluates, Western blotting analysis was performed using an anti PMCA antibody (Fig. 3h)
Figure 3h: Western blotting analysis of the partial purification of the rPMCA 2 by
Calmodulin affinity column. 1.Total cells; 2.Supernatant Homog.; 3.Membranes; 4. Pellet solub.; 5.Super. Solub.; 6.FT; 7.elution I 2mM EDTA; 8.elution II 2mM EDTA
kDa 250
1 2 3 4 5 6 7 8
148
98
kDa 250
1 2 3 4 5 6 7 8
148
98
dimer
monomer
kDa 250
148
98
dimer
monomer
kDa 250
148
98
dimer
monomer
kDa 250
148
98
A B
1 2 3 4 5 6 7 81 2 3 4 5 6 7 8
- 72 -
The level of PMCA expression was low, as expected, occurring to the lower time of
infection, but the PMCA binding to the calmodulin resin was also poorly efficient.
PMCA 2 zb seemed not to be come bound to the calmodulin, appearing almost
completely in the FT fraction. This was not expected, given the high affinity, of PMCA
2, for calmodulin. The reasons for this abnormal behavior are now investigated, thus,
we decided to investigate the binding of calmodulin to the partially purified PMCA 2
pump, estimated on PVDF membranes by exposing them to a calmodulin antibody. A
typical experiment is shown in Fig. 3i.
Figure 3i: PMCA2 is binding to Calmodulin in Sf 9 insect cells. Western blot analysis was performed with anti calmodulin antibody in different membrane fractions. 1. Membranes solubilized –Ca2+; 2. Membranes solubilized +Ca2+; 3.Membranes; 4.Membranes with 10mM EDTA; 5. Supernatant homogenized; 6. Pellet homogenized
These data, let us conclude that PMCA 2 has an extremely high calmodulin affinity,
thus, it may probably be calmodulin, saturated within the Sf9 cells and in the partially
purified state. The high affinity of PMCA 2 for calmodulin evidently complicates its
purification using Calmodulin affinity.
1.6- Construction of a Recombinant Baculovirus with a His-Tag Vector
The complete full length PMCA 2 zb and wa spliced isoforms were subcloned into a
Baculovirus shuttle fusion vector pFastBacHtB (fusion with the N-terminal 6xHis tag
and the TEV protease cleavage site). The pump was digested with Hind III from a
pcDNA3 plasmid in the case of zb splice form, and XhoI/NotI in the case of wa splice
variant.
The baculoviruses, in both cases, were produced as described above.
1 2 3 4 5 61 2 3 4 5 6
- 73 -
1.7- Time course PMCA 2 wa expression in Sf9 cells
A PMCA 2 wa expression time course was performed to analyze the time-dependent
dimerization and degradation of PMCA in Sf9 cells. Sf9 cells infected with Baculovirus
were then collected at several infection times and the protein concentration was assayed
(Table VI).
TABLE VI: PMCA2 wa expression time course protein determination
The PMCA 2 wa expression was then checked by Western blotting with an antibody
anti PMCA and with an anti Histidine Tag antibody (Fig 3j). PMCA 2 pump dimers
appears after 24 hours post infection while at 18 hours the protein was detected only by
anti His-Tag antibody. Dimerization was observed at any post infection time. A band of
90 kDa was detected by the anti His-Tag antibody but not with that against the PMCA,
possibly because this antibody recognized a protein rich in Histidine. An expression
time of 24 hours was chosen prior to the insect cells infection for the His-Tag affinity
purification
Infection times Protein concentrations
18 hours 3.8 µg/µl
24 hours 4.4 µg/µl
36 hours 7.8 µg/µl
48 hours 8.2 µg/µl
- 74 -
Figure 3j: Time course of the Sf9 insect cells infection to determine the level of the
expression of PMCA 2 wa. The membranes proteins were separated in SDS-PAGE gel and probed with 5F10 (anti PMCA) and ant His-Tag antibody 1.8- Sf 9 insect cells membrane preparation for His-Tag affinity purification
The Sf9 cell membranes were prepared and solubilized in the C12 E8 detergent as described above. The fractions infected with recombinant baculovirus PMCA 2 zb and wa were then assayed for the protein content. The protein concentrations of different membrane fractions are represented in the TABLE VII
Protein concentration Preparation
membrane Fraction PMCA 2 zb PMCA 2 wa
Membranes 9,1 µg/µl 12,5µg/µl
Membranes solub. 5,8 µg/µl 5,3 µg/µl
Pellet after solub. 7,8 µg/µl 9,1 µg/µl
TABLE VII: Sf9 membranes protein determination
The fractions were then checked by Western blotting technique with the anti PMCA 2 antibody (Fig. 3k). No degradation was detected after 24 hours of infection with either the PMCA 2 zb and wa, however, some dimerization was present.
18
24
36
48
18 24
36
48MW hs
AbAb anti HISanti HIS--TagTagAbAb anti PMCAanti PMCA
148 kDa
18
24
36
48
18 24
36
48MW hs
AbAb anti HISanti HIS--TagTagAbAb anti PMCAanti PMCA
18
24
36
48
18 24
36
48MW hs
AbAb anti HISanti HIS--TagTagAbAb anti PMCAanti PMCA
148 kDa
A B
1 2 3 1 2 3
A B
1 2 3 1 2 3
- 75 -
Figure 3k: Immunoblotting of the Sf9 membranes preparation fractions with anti
PMCA antibody A.PMCA 2 zb B.PMCA 2 wa 1.Membranes; 2.Membranes solubilized; 3.Pellet after solubilization 1.9- PMCA 2 zb and wa purification by His-Tag affinity column.
The pump splicing isoforms were solubilized with the non-ionic detergent C12 E8 at a
ratio 1:1. Purification was attended on solubilized membranes using a His-Tag spin
column. The procedure removed most unrelated proteins, but numerous contaminants
were still visible in Colloidal Coomassie gels of the imidazole eluates at different
concentrations (Fig 3l).The protein concentrations were determined by the method of
Lowry et al (TABLE VIII)
Protein concentration Preparation membrane
Fraction PMCA 2 zb PMCA 2 wa
FT 4,1 µg/µl 4,5µg/µl
Elution 25mM Imidazole 2,8 µg/µl 2,3 µg/µl
Elution 50mM Imidazole 2,5 µg/µl 2,2 µg/µl
Elution 100mM Imidazole 1,8 µg/µl 2,1 µg/µl
TABLE VIII: His-Tag affinity fractions for protein determination
Figure 3l: Partial purification of recombinant PMCA 2 by His-Tag Affinity
To verify the presence of PMCA 2 in the gel, Western blotting analysis was performed using the anti PMCA antibody (Fig. 3m)
FT E25
E50
E100
E25
E50
E100
FT
PMCA 2 PMCA 2 z/b z/b PMCA 2 PMCA 2 w/aw/a
PMCA
FT E25
E50
E100
E25
E50
E100
FT
PMCA 2 PMCA 2 z/b z/b PMCA 2 PMCA 2 w/aw/a
PMCA
- 76 -
Figure 3m: Western blotting analysis of the partial purification of the PMCA 2 by
Histidine Tag affinity column.
In order to characterize the PMCA 2 pump and facilitate the purification, we expressed
an N-terminal 6XHis-tagged version of the ATPase pump. The PMCA2 was
immunodetected as a band of about 135 kDa localized in the plasma membrane of Sf9
insect cells. The purification of the PMCA 2 zb and wa splicing pumps was
unsuccessful. The principal problem was a low level of expression and the low time of
infection in the insect cells. PMCA 2 degradation and dimerization were found after 24
hours of infection, perhaps long time of infection increases the level of expression but
the pump is degradated by calpains and other proteases that attack the C-Terminal of the
pump.
E25
E50
E100
E25
E50
E100
PMCA 2 PMCA 2 z/b z/b PMCA 2 PMCA 2 w/aw/a
E25
E50
E100
E25
E50
E100
PMCA 2 PMCA 2 z/b z/b PMCA 2 PMCA 2 w/aw/a
- 77 -
II- Biochemical study of mutants of plasma membrane calcium pump
isoform 2 causing deafness
Mutations of PMCA2 have been recently shown to induce recessively inherited
deafness (86): in the first described mouse a G283S mutation replaced a conserved G
downstream of the phospholipids binding domains of the pump (deafwaddler mouse,
dfw). A mutation next to the active center (V586M) was later found to depress pump
activity and to increase hearing loss in heterozygous patients that also carried a
homozygous mutation in cadherin 23 (CDH23) (121). A deafness inducing mutation has
then been identified in the human PMCA2 gene that replaces a conserved glycine
residue (G293S) 10 residues downstream of the dfw mutation.
Within the stereocilia, the control of Ca2+ is vital to a number of aspects of the
mechanotransduction process, e.g., the regulation of adaptation, the ability to sense the
deflection of the ciliary bundle with high sensitivity, and the breaking and regeneration
of tip links. The peculiar functions of the Ca2+ signal have in all likelihood dictated the
choice of the PMCA 2 wa isoform. Because endolymph Ca2+ is very low, it would make
sense to control it with a stereociliary pump variant that ejects Ca2+ less efficiently than
other isoforms.
In the second part of this work, the activity of the recombinant wa variant of the
PMCA2 pump was thus compared in a Baculovirus expression system with the zb
variant of the pump and with the dfw pump and the pump carrying the G293S human
mutation. The splicing variants of the pump and the mutations have been studied
biochemically by a coupled enzyme assay.
A novel mutation (Oblivion) was also studied; this mutation is localized in the sixth
transmembrane domain of the PMCA 2 wa pump where a Serine is replaced by a
Phenylalanine at position 877. The Oblivion mice present ataxia, loss of balance and
progressive loss of the auditory function.
- 78 -
1- Construction and expression of the PMCA2 wa mutants
The experiments were particularly aimed at understanding the biochemical properties of
the mutants, of the wa splice variant of the PMCA 2 pump which is the resident isoform
of the stereocilia. The mutagenesis was performed by site-directed mutagenesis and the
cassettes containing the mutations were sequenced to completion in at least one
direction. The mutated PMCA 2 pumps were expressed in Sf9 insect cells using the
Baculovirus System. The mutants, the wild type stereociliary pump (wa) and the PMCA
2 zb pump used as a control were expressed at about the same levels, indicating that the
mutations had not increased the propensity of the pump to become proteolyzed.
In crude membranes of Sf9 cells, a band migrating with a molecular mass of 135 kDa
was recognized by the PMCA2 antibody. (Fig. 3n)
Figure 3n: Transient expression of the mutated PMCA 2 wa pumps in Sf9 insect
cells. A and B, 10 µg of crude membrane proteins prepared by freeze and thaw method, were separated by SDS-PAGE, transferred to PVDF membranes and blotting with antibody specific for PMCA 2 pump. In the panel A the antibody recognizes the dfw mutant (G283S) and the human mutation (G293S). In panel B Oblivion mutation was also stained by PMCA 2 antibody. PMCA 2 zb was used as a control.
1.1-Determination of the Ca2+
transport activities of microsomes from PMCA 2 zb,
PMCA 2 wa and of mutated pumps of the wa variant.
Microsomal membranes isolated from PMCA 2 zb, PMCA 2 wa, PMCA 2 G283S,
G293S and Oblivion infected cells were assayed in the presence of thapsigargin and
oligomycin to inhibit the activity of the endogenous endoplasmic reticulum Ca2+ pump
and the mitochondrial ATP mediated Ca 2+ uptake system (143). In Sf9 cells
membranes expressing the PMCA 2 zb, the Ca2+ ATPase was higher than in those of
cells expressing the PMCA 2 wa pump. The specific activity of the G283S mice pump
G293S
PL
32
G283S
G283S
G293S
PMCA 2 z/b
A
G293S
PL
32
G283S
G293S
PL
32
G283S
PL
32
G283S
G283S
G293S
PMCA 2 z/b
A
76
S877FPMCA 2 z/b
OBLIVION
B
76
S877FPMCA 2 z/b
OBLIVION
76
S877F
76
S877FPMCA 2 z/b
OBLIVION
B
- 79 -
mutant and G293S human mutant were almost the same but were lower than that of the
wild type wa pump. The concentration level of the protein expression was verified by
Western blotting. (Fig. 3o)
Figure 3o: Comparison of Ca
2+ uptake activities of microsomes isolated from
PMCA 2 zb, PMCA 2 wa, PMCA 2 wa G283S, PMCA 2 wa G293S and PMCA 2
wa Oblivion infected Sf9 cells. A. The Ca2+ ATPase activity was measured spectrophotometrically using coupled enzyme assay. The medium contained 120 mM KCl, 60 mM Hepes, 1mM MgCl2, 2.5 mM ATP, 0.2 mM NADH, 0.5 mM PK/PL, 1 unit of LDH and 20 mM EGTA. The assay was performed in presence of oligomycin and thapsigargin. B. Western blotting assay using PMCA 2 antibody
The calcium uptake activity (measured as ATP splitting) was calculated as ∆
Absorbance / ∆ time from each curve and it is represented in the Figure 3p. The error
bar represents the standard errors of the mean.
zb wa 283 293 OBL
0 100 200 300 400 500 600 700
0,5
0,6
0,7
0,8
0,9
1,0
1,1
Ab
so
rban
ce (
340 n
m)
Time (sec)
zb
waOBL
wa283
wa293
wa
A
Bzb wa 283 293 OBLzb wa 283 293 OBL
0 100 200 300 400 500 600 700
0,5
0,6
0,7
0,8
0,9
1,0
1,1
Ab
so
rban
ce (
340 n
m)
Time (sec)
zb
waOBL
wa283
wa293
wa
A
B
- 80 -
Figure 3p: The Ca2+
uptake activity of the PMCA2 zb, wa and mutants pumps.
The Ca2+ activity of the PMCA 2 zb pump was almost 2 fold higher than PMCA2 wa. The wa 283 and wa 293 had lower activity than wild type pump. At the 0,05 level the means are significantly different.
These data suggest that, the calcium transport activity defect with respect to the wild
type wa variant was more pronounced in the dfw mutant (G283S) than in the G293S
(human) mutant. The PMCA 2 of the Oblivion mouse was slightly more active than the
pumps bearing the 283 and 293 mutations but still much lower that of the wild type.
1.3- Study of a novel mutation near to the ATP binding domain of the PMCA 2
pump.
Another mutation (2075G A) results in the substitution of a methionine for a valine at
amino acid position 586 (V586M) in the T4–T5 intracellular catalytic loop of PMCA2.
Molecular modeling of V586M based on the template of the three-dimensional structure
of the closely related sarcoplasmic reticulum calcium pump predicts that substitution
with the sterically larger methionine side chain would distort packing underneath the
ATP-binding interface or increase its projection from the external solvent-exposed
surface of the nucleotide-binding domain. The valine residue at position 586 is
completely conserved among mouse, rat, and fish PMCA2 orthologues, and either
valine or a conservatively substituted residue (isoleucine) are present at this position in
zb OBLIVION wa283 wa293 wa0,00
0,02
0,04
0,06
0,08
0,10
0,12
0,14
Acti
vit
y
zb OBLIVION wa283 wa293 wa0,00
0,02
0,04
0,06
0,08
0,10
0,12
0,14
Acti
vit
y
- 81 -
all known PMCA1 and PMCA3 amino acid sequences. The V586M mutation was
detected in a mouse, called Tommy, that were affected by sensorineural hearing loss,
suggesting that this variant may modify the severity of sensorineural hearing loss caused
by a variety of factors.
Crude membranes from CHO cells transfected with a cDNA encoding the 2 zb, PMCA2
wa and Tommy (V586M) variant were prepared and separated by SDS-polyacrylamide
gel. The proteins were analyzed by Western blotting and stained by Coomassie Brilliant
Blue. (Fig. 3q)
Figure 3q: Coomasie and Western Blotting analysis of PMCA 2 zb, wa and Tommy
pumps
Microsomal membranes isolated from CHO cells transfected with PMCA 2 zb, PMCA2
wa and Tommy (V586M) mutants were assayed in the presence of thapsigargin and
oligomycin. PMCA 2 zb had higher reactivity to Ca2+ than PMCA 2 wa. When
expressed in CHO cells, PMCA2 Tommy variant had approximately 25 percent of the
calcium stimulated ATPase activity of the wild-type PMCA 2 wa. (Fig.3r)
Zb wa
To
mm
y
Zb wa
To
mm
y
- 82 -
Figure 3r: The Ca2+
uptake activity of the PMCA2 zb, wa and Tommy mutant.
The calcium ATPase activity of PMCA2 zb, PMCA2 wa and Tommy was determined in the absence of calmodulin. Three ATPase experiments were performed for each expressed protein from three microsomal preparations of three different expressed-protein preparations. Representative data are shown from one of the three experiments.
0 100 200 300 400 500 600 700
0,4
0,5
0,6
0,7
0,8
0,9
1,0
1,1
Ab
so
rban
ce (
340n
m)
TIME (sec)
PMCA2zb
PMCA2wa TOMMY
0 100 200 300 400 500 600 700
0,4
0,5
0,6
0,7
0,8
0,9
1,0
1,1
Ab
so
rban
ce (
340n
m)
TIME (sec)
PMCA2zb
PMCA2wa TOMMY
- 83 -
III- Functional properties of the PMCA 2 wa splice variant.
The PMCA2 isoforms (the products of gene ATP2B2) are transcribed in a limited
number of tissues, chiefly, brain and skeletal muscle (78, 95,100). Their sequences have
been deduced independently by different groups (78, 79, 82), and their human gene has
been mapped to chromosome 3 (3p25-26) (78, 79, 80), some 200-500 kilobase pairs
away from the locus containing the gene associated with the von Hippel-Lindau disease
(79). Specific brain regions, e.g. the Purkinje’s cells layer, contain high levels of PMCA
2 transcripts (87).
During the isolation of human clones for the PMCA2 pump an alternative splicing
process was discovered in the N-terminal half of the protein (82). The C-terminal
splicing of PMCA transcripts had been described previously (93). The 1 to 3 exons will
insert in the transcript (95) (splicing site A). The insertion occurred in the region
corresponding to the mid portion of the cytosolic loop between transmembrane domains
2 and 3. A region of 65 amino acids just downstream of this splicing site was suggested
to modulate the activation of the PMCA pump by the calmodulin binding domain (46),
and a region upstream of it was proposed to interact with activating acidic
phospholipids (51).
The aim of this chapter was the study of the properties of the PMCA 2 and its splicing
products, represented in the Figure 3s
Figure 3s: Schematic representation of the PMCA 2 splicing variants
PMCA 2 zb
Splicing site A Splicing site C
PMCA 2 wa
PMCA 2 wb
PMCA 2 zb
Splicing site A Splicing site C
PMCA 2 wa
PMCA 2 wb
PMCA 2 zb
Splicing site A Splicing site C
PMCA 2 wa
PMCA 2 wb
- 84 -
1- Functional studies of PMCA wa. Unusual variant of PMCA 2 pump.
1.1- Expression of PMCA 2 splicing variants in CHO cells
To study the Ca2+ transporting activity of the PMCA 2 splice variants zb, wb, wa and
that of the mutant Tommy PMCA wa pumps, the full-length clones were overexpressed
in CHO cells. Crude membranes from cells transfected with cDNA encoding the
splicing variants and the mutant were prepared and their protein (1 µg) separated by a
7.5% SDS-polyacrylamide gel. Fig. 3t shows a Western blot using the 5F10 antibody
and gene reporter GFP monoclonal antibodies to visualize the variants and the mutant.
The estimated molecular weights corresponded to the expected molecular masses of
each isoform based on their protein sequences. The level of expression was about the
same for all four isoforms.
Figure 3t: Immunoblot of microsomes prepared from CHO cells transfected with
plasmid encoding GFP and pumps proteins
1.2- Determination of Ca2+
Transport Activities of Microsomes from transfected
CHO Cells with PMCA 2 zb, PMCA 2 wb, PMCA 2 wa and PMCA 2 wa Tommy.
Microsomal membranes isolated from cells transfected with PMCA 2 zb, PMCA 2 wb,
PMCA 2 wa and PMCA 2 wa (Tommy) were assayed in the presence of thapsigargin
and oligomycin to inhibit the activity of the endogenous endoplasmic reticulum Ca2+
pump and the ATP linked mitochondrial Ca2+ uptake as described above. Figure 3u
shows Ca2+ uptake by the microsomal in the absence of calmodulin. Both PMCA 2 zb
WB: anti PMCA
WB: anti GFP
wb
To
mm
y
zb
wa
WB: anti PMCA
WB: anti GFP
wb
To
mm
y
zb
wa
WB: anti PMCA
WB: anti GFP
wb
To
mm
y
zb
wa
- 85 -
and PMCA 2 wb had higher basal activity than PMCA 2 wa. PMCA 2 wa (Tommy)
only had about 25 per cent of the activity of PMCA 2 wa (see previous section).
Figure 3u: Comparison of Ca2+
activities of microsomes isolated from PMCA 2 zb,
PMCA 2 wb, PMCA 2 wa and PMCA 2 wa Tommy pumps transfected CHO cells.
A. Activity of recombinant PMCA2 isoforms and of the wa isoforms in CHO cells. CHO cells were transiently transfected with the PMCA2 variants. Membrane vesicles were preincubated at 37 °C and Ca2+ uptake was initiated by the addition of ATP. The histograms show the means +/- SD of the activity of the pumps. The traces are
representative of at least three experiments with different membranes preparations. At the 0,05 level the means are significantly different.
1.3- Determination of Ca2+
Transport Activities of Microsomes from CHO Cells
transfected with PMCA 2 zb, PMCA 2 wb, PMCA 2 wa and PMCA 2 wa Tommy
transfected. Calmodulin activation
The calmodulin sensitivity of each isoform was observed by measuring its Ca2+
transport activities at a fixed Ca2+ concentration in the presence of the 200 nM
calmodulin (Fig. 3v). PMCA2 zb and wb showed the highest response to calmodulin
i.e., the activity of these isoforms was over eight times higher than of PMCA 2 wa.
PMCA2 wa had lower affinity for calmodulin than PMCA2 zb and wb, but had still
higher affinity than the corresponding Tommy mutant.
0 100 200 300 400 500 600 700
0,5
0,6
0,7
0,8
0,9
1,0
1,1
1,2
Ab
so
rban
ce
(340n
m)
Time (sec)
PMCA2zb
PMCA2wb
PMCA2wa
TOMMY
0 100 200 300 400 500 600 700
0,5
0,6
0,7
0,8
0,9
1,0
1,1
1,2
Ab
so
rban
ce
(340n
m)
Time (sec)
PMCA2zb
PMCA2wb
PMCA2wa
TOMMY
zb wb wa Tommy0,00
0,02
0,04
0,06
0,08
0,10
0,12
0,14
Activity
zb wb wa Tommy0,00
0,02
0,04
0,06
0,08
0,10
0,12
0,14
Activity
A B
- 86 -
Figure 4v: Fig. 2. PMCA 2 zb and wb are more sensitive to calmodulin stimulation
than PMCA wa. A.The calmodulin dependence of Ca2+ uptake by microsomal vesicles isolated from CHO cells transfected with PMCA 2 zb (green curve), PMCA 2 wb (light blue), PMCA 2 wa
(blue) and Tommy (orange) is shown. Membrane vesicles were preincubated at 37 °C with 200 nM of calmodulin and Ca2+ uptake was initiated by the addition of ATP. B. The histograms show the means +/- SD of the activity of the pumps. At the 0,05 level the means are significantly different
1.4- Determination of Ca2+
Transport Activities of Microsomes from CHO Cells
transfected with PMCA 2 zb, PMCA 2 wb, PMCA 2 wa and PMCA 2 wa Tommy.
Activation by Phosphatidylserine
Acidic phospholipids bind to the PMCA protein and activate it by interacting with two
domains, the calmodulin binding domain and a sequence close to splicing site A (51).
The splicing event that occurs at site A affects the sequence immediately 5' to the region
that encodes one of the phospholipid-sensitive portions of the PMCA .Acidic
phospholipids, in particular polyphosphoinositides, are potent activators of the PMCA.
As mentioned, splice site A is situated between the phospholipid binding region and a
sequence further downstream that is involved in an intramolecular inhibitory interaction
with the C-terminal calmodulin binding domain
Data showing that the Ca2+-calmodulin-stimulated pump (or the truncated a variant of
the pump lacking the COOH-terminal autoinhibitory, calmodulin binding domain) could
be further stimulated by phospholipids indicated that a region(s) other than the COOH-
terminal domain which also binds acidic phospholipids must also be involved in the
lipid stimulation. The mechanism by which different phospholipids activate the PMCAs
is not known. However, given the location of the lipid-binding sequences in the pump,
0 100 200 300 400 500 600 700
0,0
0,2
0,4
0,6
0,8
1,0
1,2
Ab
so
rban
ce (
340n
m)
Time (sec)
PMCAzb
PMCAwb
PMCAwa
TOMMY
A B
zb wb wa Tommy
0,0
0,5
1,0
1,5
2,0
2,5
Activity
0 100 200 300 400 500 600 700
0,0
0,2
0,4
0,6
0,8
1,0
1,2
Ab
so
rban
ce (
340n
m)
Time (sec)
PMCAzb
PMCAwb
PMCAwa
TOMMY
0 100 200 300 400 500 600 700
0,0
0,2
0,4
0,6
0,8
1,0
1,2
Ab
so
rban
ce (
340n
m)
Time (sec)
PMCAzb
PMCAwb
PMCAwa
TOMMY
A B
zb wb wa Tommy
0,0
0,5
1,0
1,5
2,0
2,5
Activity
B
zb wb wa Tommy
0,0
0,5
1,0
1,5
2,0
2,5
Activity
- 87 -
one speculated that the interaction of acidic phospholipids (presumably via their charged
headgroups) with the calmodulin binding domain leads to some structural rearrangement
that weakens the autoinhibitory intramolecular interactions formed by the COOH-
terminal tail
The Ca2+-dependent ATPase associated with CHO membranes of cells expressing the
PMCA 2 zb, wb, wa and wa Tommy was stimulated by phosphatidyl serine. The PMCA
2 zb and wb had the same response to this acidic phospholipids implying that the inserts
next to the N-terminal phospholipids binding domain (variant w) was not disturbed by
the splicing insert. The PMCA wa was absolutely not stimulated by phosphatidyl serine
(Fig. 4w): apparently, then, no activation by acidic phospholipids can occur when the
C-terminal binding site is undisturbed. The response of the PMCA 2 b pumps to PS was
4 fold higher than that of the wa variant.
Figure 4w: Stimulation of the Ca 2+
-ATPase by acidic phospholipid A. The PS dependence of Ca2+ uptake by microsomal vesicles isolated from CHO cells transfected
with PMCA 2 zb (green curve), PMCA 2 wb (light blue), PMCA 2 wa (blue) and
Tommy (orange) is shown. Membrane vesicles were preincubated at 37 °C with 25 µM PS and Ca2+ uptake was initiated by the addition of ATP. B. The histograms show the
means +/- SD of the activity of the pumps. At the 0,05 level the means are significantly different
0 100 200 300 400 500 600 700
0,0
0,2
0,4
0,6
0,8
1,0
1,2
Ab
so
rba
nce (
340n
m)
Time (sec)
PMCA2zb
PMCA2wb PMCA2wa
TOMMY
zb wb wa Tommy-0,1
0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
Activi
ty
A B
- 88 -
IV- Structural model of PMCA 2 and a biochemical study of the effect
of a Met265Ala mutation in one of the domain that mediate the auto-
inhibition of the pump
1- Comparative model between SERCA and PMCA 2
To investigate whether CaM indeed removes its binding domain from the cytosolic
loops of the pump, we decide to introduce a mutation in one of the two sites that interact
with the CaM binding domain in the cytosolic loops of the pump. Prior to this, the
PMCA pump has been modeled to decide which residue to mutate. The sequence
alignment presented in Fig. 3x illustrates the degree residues conservation among the
PMCAs and the SERCA pump template sequences. Both sequences align with
relatively low conservation sowing gaps and insertions, among which the largest
correspond to the splicing site A. Also as expected, the phosphorylation site is well
conserved while the residues involved in the recognition of Thapsigargin (144) by the
SERCA pump are not conserved in the PMCA pumps. In agreement with the fact that
this compound does not inhibit PMCA pumps.
Based on this sequence alignment, we have constructed a comparative modeling-based
3D structure of the PMCA. The model was built with Modeller 8.0 with the automatic
loop refinement option Among the different structures available for the SERCA pump
we chose the calcium bound form (E1 state) (23) (PDB entry code:1SU4) since it
presents a more open and less structured conformation near the gap corresponding to the
insertion at splicing site A in the PMCA pump, facilitating the modelling (Fig. 3y)
- 89 -
Figure 3x. Sequence Alignment between SERCA and PMCA 2 pumps. Residues are shaded by conservation. Red rectangle indicates the splicing site A insertion. Yellow rectangle indicates the conserved phosphorylation site. Green circles highlight residues involved in TG recognition in SERCA pump not conserved in PMCA pumps. The sequences of the splicing variants wa and zb are included to illustrate the differences in the insertion length at Site A.
- 90 -
Figure 3y: Choice of the Structural Template. Root mean square deviation fit between SERCA pump in the calcium free E2 (a), and calcium bound, E1 states (b). Cartoons are colored by B-factor. Blue regions are better defined than red regions. Residues shown in lines display a temperature factor higher than 120 Å2 (poor structural determination). Notice the higher exposition of the segment corresponding to the splicing site A in the E1 structure
To gain structural insights into the mechanism of auto inhibition of the PMCAs we
constructed a 3D model of the zb variant of this protein using the above-described
alignment. The theoretical model selected superimposed on its structural template with
a root mean square deviation of 0,22 nm (Fig. 3z). Although the relatively low identity
degree (nearly 30%) did not warrant structural reliability up to atomic resolution, the
theoretical model revealed some interesting features. The amino acids corresponding to
the site of insertion of the splicing site A remain at less than 1 nm from the membrane
plane allowing the stretch of residues from KKA to KLA to get in direct contact with
the inner leaflet of the plasma membrane. This may suggest that a direct interaction that
might pull down the N domain of PMCA hampering its capacity to bind ATP. This
block being then released upon addition of acidic phospholipids
Digestion of the peptides resulting from cross-linking experiments have identified
peptides DK to VV (region from 206-271 in the first cytoplasmatic loop) and AL to TD
Splice site A
Insertion
a) b)
Splice site A
Insertion
a) b)
- 91 -
(region from 537 and 544 in the second cytoplasmatic loop) to be belong to the binding
site of the C-terminal calmodulin binding domain (46). The reactant used had a marked
reactivity with sulphur atoms and four sulphur containing residues are indeed found in
those peptides (Met257, 265 and 271 in the first segment and Cys571 in the second
segment). While, Cys571 is likely to be part of the binding site, the length of the first
pinpointed peptide makes it difficult to rationalize this information. Therefore we used
the structural information provided by our model to refine this information. Among
these residues Met257 resulted buried within the N domain. Furthermore, it is not
conserved in isoform 1 of PMCA strongly suggesting that it may not participate of the
binding site. Of the remaining two, only Met265 is oriented towards the putative
binding site (Fig 3z). Therefore, it can be expected that a mutation of this residue may
impair the capability of the pump to bind the calmodulin binding domain.
Figure 3z: Structural model of the PMCA2 pump.
a) Superposition of the selected model of PMCA2 on the structure of the SERCA pump in E1 state (PDB entry code 1SU4). b) Model of PMCA2 pump. The loopy region were splicing inserts occur is marked with red tube and corresponds to the less structurally resolved region of the SERCA pump. c) The segments near to the auto-inhibitory binding site identified by cross linking experiments are shown in sticks. The four putative sulphur containing residues (Cys571, Met257, Met265 and Met271) within these segments are highlighted in yellow. Notice that the residue corresponding to Cys571 in the SERCA pump lies within the ATP binding site. Met257 and 271 are not solvent exposed. Furthermore, Met257 is not conserved in PMCA1 as shown the alignment. The green circle indicates the putative binding site of the calmodulin binding peptide.
To verify this hypothesis, Met265 was mutated to Ala in the PMCA 2 zb pump. The
mutated PMCA 2 was expressed in parallel with wild type pump in CHO cells. The
expression of the PMCA protein in CHO membranes was detected by SDS-PAGE and
Western blot (Fig. 3za). Antibody 5F10 recognized a single band at 135 kDa indicating
that the full-length polypeptide of the recombinant wild type protein was successfully
- 92 -
expressed. Mutant M265A was expressed at a level similar to that of the wild type
protein.
Figure 3za: Expression of PMCA wild type and M265A mutant. 1 µg of total CHO membranes protein was applied on each lane of an SDS-10% polyacrylamide gel stained with Coomasie Brilliant Blue and subjected to immunoblot analysis with 5F10 monoclonal antibody. 1.1- Functional State of the PMCA 2 zb and M265A mutant.
The activity of the PMCA 2 zb and M265A pumps expressed in transfected CHO cells
was evaluated as a function of Ca2+ in both the presence and absence of calmodulin.
Both conditions, the M265A mutant had a response to calmodulin which was the same
as that of the wild type when the concentration of calmodulin was 200 nM. The kinetic
parameters of PMCA 2 zb and M265A are showed in Fig. 3zb.
MW zb M265A
zb M265A
135 kDa
MW zb M265A
zb M265A
MW zb M265A
zb M265A
135 kDa
- 93 -
Figure 3zb: The rate of ATP hydrolysis by PMCA 2 zb and M265A enzymes. The Ca 2+ ATPase activity of PMCA membranes was measured at 37 °C as described under "Materials and Methods" in a medium containing 20 mM HEPES, pH 7.00 at 37 °C, 120 mM KCl, 4 mM MgCl2, 0.5 mM EGTA, 3 mM , and enough CaCl2 to give 10 µM Ca2+. A. no calmodulin B. 200nM calmodulin. This data is a representative curve of 3 experiments with 3 different preparations.
The PMCA 2 zb and M265A mutant had the same response to phosphatidyl serine. (Fig.
3zc)
Figure 3 zc: Stimulation of the Ca 2+
ATPase by phosphatidyl serine The Ca2+ uptake by microsomal vesicles isolated from CHO cells transfected with PMCA 2 zb (green curve), PMCA 2 M265A (orange) are shown. Membrane vesicles were preincubated at 37 °C with 25 µM PS and Ca2+ uptake was initiated by the addition of ATP. No differences were found between wild type and mutated pump.
0 100 200 300 400
0,3
0,4
0,5
0,6
0,7
0,8A
bso
rban
ce (
nm
)
Time (sec)
zb
M265A
200nM CaM
0 20 40 60 80 100 120 140 160 180
0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
Ab
so
rban
ce (
nm
)
Time (sec)
zb
M265A
200nM CaM
0 20 40 60 80 100 120 140 160 180
0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
Ab
so
rban
ce (
nm
)
Time (sec)
zb
M265A
A B
0 100 200 300 400
0,3
0,4
0,5
0,6
0,7
0,8A
bso
rban
ce (
nm
)
Time (sec)
zb
M265A
200nM CaM
0 20 40 60 80 100 120 140 160 180
0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
Ab
so
rban
ce (
nm
)
Time (sec)
zb
M265A
200nM CaM
0 20 40 60 80 100 120 140 160 180
0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
Ab
so
rban
ce (
nm
)
Time (sec)
zb
M265A
A B
0 50 100 150 200 250
0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
Ab
so
rban
ce (
nm
)
Time (sec)
zbPS
M265APS
0 50 100 150 200 250
0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
Ab
so
rban
ce (
nm
)
Time (sec)
zbPS
M265APS
- 94 -
We then tested the effect of the mutation at Met265 on the ability of calmodulin to
activate the pump. The Ca2+ transport activity was evaluated as a function of the
calmodulin concentration. To reach steady state or near steady state conditions, the
membranes were preincubated with calmodulin for 5-10 min at 37 °C. The results
showed that when the concentration of calmodulin was decreased the activation became
higher for M265A with respect to the wild type pump. (Fig. 3zd and ze). At 1, 5, 15 nM
calmodulin the activity of the mutant was invariabily higher than that of the wild type
pump. Evidently, the inhibition of Met 265 decreased the affinity of the C-terminal
autoinhibitory domain, making it easier for calmodulin to remove it from binding site.
Figure 3zd: Calmodulin dependence of the rate of ATP hydrolysis by the PMCA 2
zb and M265A. ATP hydrolysis was measured as described under "Materials and Methods" at a constant free Ca2+ of 10 µM. Calmodulin concentrations used were 1, 5 and 15 nM. Three differents experiments were performed with different membranes preparations
0 50 100 150 200 250
0,3
0,4
0,5
0,6
0,7
0,8
Ab
so
rban
ce (
nm
)
Time (sec)
zb M265A
0 20 40 60 80 100 120 140 160 180
0,25
0,30
0,35
0,40
0,45
0,50
0,55
0,60
0,65
0,70
Ab
so
rban
ce (
340 n
m)
TIME (sec)
zb
M265A
0 20 40 60 80 100 120 140 160 180
0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
Ab
so
rban
ce (
340 n
m)
Time (sec)
zb
M265A
A
B
C
0 50 100 150 200 250
0,3
0,4
0,5
0,6
0,7
0,8
Ab
so
rban
ce (
nm
)
Time (sec)
zb M265A
0 20 40 60 80 100 120 140 160 180
0,25
0,30
0,35
0,40
0,45
0,50
0,55
0,60
0,65
0,70
Ab
so
rban
ce (
340 n
m)
TIME (sec)
zb
M265A
0 20 40 60 80 100 120 140 160 180
0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
Ab
so
rban
ce (
340 n
m)
Time (sec)
zb
M265A
A
B
C
- 95 -
Figure 3 ze: The activity of the PMCA 2 zb and M265A at different calmodulin
concentrations. The histograms show the means +/- SD of the activity of the pumps. At the 0,05 level the means are significantly different
zbnoCaM M265AnoCaM zbCaM1nM M265A1nMCaM zb5nmCaM M265A5nmCaM zb15nmCaM M265A15nmCa zb200nmCaM M265A200nmC
0,0
0,2
0,4
0,6
0,8
1,0
1,2
1,4
1,6
Acti
vit
y
- 96 -
DISCUSSION
Calcium ions are the main second messengers in all types of eukaryotic cells, in which
the free Ca2+ concentration in the cytosol is maintained at a very low level (50–150
nM). Fluctuations around this are essential for the control of normal cellular activities
and are closely connected with a number of events, e.g., the origin and development of
cells, mitotic activity, immune response, muscle contraction, endo- and exocytosis, or
modulation of neuronal cells processes also through the control of neurotransmitter
release. The precise regulation of the Ca2+ homeostasis in cells is a result of the
concerted functioning of transporters located in the plasma membrane, systems
operating in cell organelles, i.e., the endo/sarcoplasmic reticulum, the mitochondria, the
nucleus, and a number of calcium binding proteins. Calcium efflux from excitable cells
occurs through two main systems, an electrochemically driven Na+/Ca2+ exchanger with
low Ca2+ affinity, and a plasma membrane Ca2+ -ATPase, with much higher Ca2+
affinity (Guerini, 1998). The transport capacity of the fully activated Na+/Ca2+
exchanger is more than 10 times greater than that of the plasma membrane Ca2+-
ATPase. In nonexcitable cells the calcium pump is the sole system responsible for the
extrusion of calcium ions given its high Ca2+ affinity, the plasma membrane Ca2+-
ATPase is considered as the fine tuner of cytosolic calcium ion concentration, and its
important role is also reflected in its isoform-specific expression among different cell
types.
As found with other ion pumps, PMCAs form a multigene family. The basic plasma
membrane calcium pump isoforms are encoded by 4 different genes. Due to alternative
mRNA splicing more than 30 variants of the pump exist (Guerini, 1998). Based on
functional studies it has been shown that these isoforms and variants differ in regulatory
properties.
PMCA isoforms 1 and 4 are, in general, expressed in most tissues, whereas PMCA
isoforms 2 and 3 are expressed in a much more restricted manner and, in the adult,
predominantly in the brain and striated muscle. Within the brain, PMCA2 is primarily
expressed within specialized cell types such as the cerebellar Purkinje cells (Stauffer et
al., 1997) and cochlear hair cells (Furuta et al., 1998; Street et al., 1998), where the
pump is specifically expressed in the stereocilia of hair cells. Isoform 2 of the pump has
properties that distinguish it from other PMCA isoforms: When tested in the cellular
environment, it is two to three times more active than the two ubiquitous pumps in
- 97 -
pumping Ca2+ out of the cell (Brini et al., 2003). In the isolated state, it shows very high
calmodulin affinity (Kd, 2–4 nM) (Elwess et al., 1997); i.e., it becomes fully activated
under conditions (e.g., calmodulin and/or Ca2+ concentration) that would activate only
poorly isoforms 1 and 4.
Alternative splicing is peculiarly complex in PMCA2 because it involves the insertion
of up to three novel exons at site A and of two at site C. The A-site insertions are in-
frame, creating variant w when three exons are inserted (the normal variant without site-
A inserts is termed z). The insert at site C creates instead a novel stop codon, leading to
the truncation of the pump (variants a and b being the normal full-length pump). The
site-C insertions eliminate approximately half of the calmodulin binding domain; those
at site A occur next to a domain that binds activatory acidic phospholipids, which,
however, also bind to the calmodulin binding domain. The site-A insertion could thus
impair the activation by acidic phospholipids, particularly in the C-terminally truncated
pump variants, in which the C-terminal phospholipid binding domain also is
compromised.
Of interest to the topic of this thesis is how alternative splicing influences the regulation
of the PMCAs, primarily their activation by calmodulin. For instance, the stereocilia of
the outer hair cells (OHCs) contain a splice variant (wa) that has high basal Ca2+
ejection activity but fails to respond with a rapid activation to the sudden arrival of a
Ca2+ load; these properties evidently satisfy the Ca2+ homeostasis demands of both the
endolymph and of the stereocilia. Even if Ca2+ in the endolymph is very low it, is
considerably higher than that necessary for the integrity of the tip links. Within the
stereocilia, the control of Ca2+ is vital to a number of aspects of the
mechanotransduction process, e.g., the regulation of adaptation (Gillespie et al 2004),
the ability to sense the deflection of the ciliary bundle with high sensitivity, and the
breaking and regeneration of tip links (Zhao et al., 1996).As mentioned, these peculiar
functions of the Ca2+ signal have in all likelihood dictated the choice of the PMCA 2 wa
isoform for the stereocilia. PMCA 2 zb has a faster calcium-activation by Ca2+ pulse
than PMCA 2 wa. The lower response of the wa variant to large calcium loads makes it
unable to eject them rapidly. This may be required to satisfy the calcium homeostasis
demands of stereocilia, which contains much less calcium than normal extracellular
fluids.
Because resting stereociliary Ca2+ is very low, it makes sense to control it with a pump
variant that decreases Ca2+ to lower concentrations than other isoforms (Elwess et al.,
- 98 -
1997), even if it is insensitive to calmodulin. The relative insensitivity of the wa pump
to calmodulin makes good sense also, given its very high concentration in the stereocilia
(Walzer et al., 1996), it would produce permanent maximal activation of pump isoforms
normally sensitive to it. On the other hand, any variation in the activity of the pump
could conceivably have important consequence on endolymphal Ca2+.
Another important aspect of the PMCA regulation is the activation by acidic
phospholipids. Acidic phospholipids, polyphosphoinositides in particular, are the most
potent stimulators of the PMCA (Niggli et al., 1981). They reduce the Km for Ca2+ (Km
Ca2+) of the enzyme to 0.3 µM as compared with 0.4 – 0.7 µM in the case of calmodulin
stimulation (Penniston et al., 1998). The activation by acidic phospholipids renders the
PMCA insensitive to calmodulin activation, i.e., Ca2+-calmodulin and phospholipids are
not additive although acidic phospholipids can further reduce the Km for Ca2+ of the
calmodulin-stimulated pump (Wang et al., 1991). Data showing that the fully
calmodulin-stimulated pump (or of a truncated pump lacking the COOH-terminal
autoinhibitory calmodulin binding domain) could be further stimulated by
phospholipids indicated that a region(s) other than the COOH-terminal domain must
also be involved in the lipid stimulation (Enyedi et al., 1987). This was to be expected,
as work with different proteolytic fragments of the PMCA and with synthetic peptides
revealed that there are indeed two separate phospholipid binding regions in the pump:
one corresponding to the C-terminal calmodulin binding domain and the other situated
in the first cytosolic loop immediately preceding the third membrane-spanning domain
(Zvaritch et al., 1990). However, the mechanism by which different phospholipids
activate the PMCAs is not known. Given the location of the lipid-binding sequences in
the pump, one may speculate that the interaction of acidic phospholipids (presumably
via their charged headgroups) with the calmodulin binding domain leads to some
structural rearrangement that weakens the autoinhibitory intramolecular interactions
formed by the C-terminal tail. How phospholipid interactions with the first cytosolic
loop further stimulate PMCA activity in a fully calmodulin-activated pump is less
obvious. Structural rearrangements of the transduction and catalytic domains could be
invoked that may facilitate access of Ca2+ to its high-affinity transport sites and/or
positively influence a rate-limiting step in the catalytic cycle. Under stimulation by
phosphatidyl serine (PS), the PMCA 2 zb and wb, transfected in CHO cells, have the
same response to PS, even if the wb variant has an insert close to the end terminal of the
phospholipid binding domain (PL-BD), which could be expected, interfere with PL
- 99 -
binding domain nearby. Of the two PL binding domains that in the calmodulin binding
domain is apparently more critical in the activation by acidic phospholipids. This is
directly supported by the finding that the wa variant, which has lost, at least in part, the
PL-BD in the CaM binding sequence is totally unresponsive to phospholipids.
A recent finding on PMCA 2 relates its defect to hearing deafness. These uncertainties
of mechanism have prompted the experiments described in this thesis:
A variety of recessive mutations of cadherin 23 (Cdh23) cause profound deafness and
vestibular dysfunction in homozygous mice,( Di Palma et al., 2001 ) whereas another
allele of Cdh23, called ahl, underlies less severe, age-related hearing loss in many
inbred mouse strains (Noben-Trauth et al., 2003). The severity of this age-related
hearing loss is significantly increased by heterozygosity for the dfw2j deafwaddler allele
of Atp2b2, which encodes PMCA2, the predominant PMCA of hair bundles. This
interaction has been attributed to a reduction in PMCA2 activity that would result in a
decrease in extracellular (endolymphal) calcium concentrations around hair bundles,
where calcium-dependent, cadherin-mediated adhesion is thought to occur. The 2075G-
A mutation is predicted to result in the substitution of methionine for valine at amino
acid position 586 (V586M) in the T4–T5 intracellular catalytic loop of PMCA2. The
V586M mutation was detected in a mouse, called Tommy, which was affected by
sensorineural hearing loss, suggesting that this pump variant may modify the severity of
sensorineural hearing loss caused by a variety of factors. The valine residue at position
586 is completely conserved among mouse, rat, and fish PMCA2 orthologues, and
either valine or a conservatively substituted residue (isoleucine) is present at this
position in all known PMCA1 and PMCA3 aminoacid sequences. When expressed as a
recombinant baculovirus protein in Sf9 cells, PMCA2 wa V586M has only
approximately 50 percent of the calcium ATPase activity of wild-type PMCA 2 wa.
Other calcium transport activity defect with respect to the wild type wa variant have
been found in mice, for instance a G283S mutation in deafwaddler mouse and in
humans (G293S). The defect is more pronounced in the dfw mutant than in the G293S
(human) mutant when were analyzed in Baculovirus expression system. The PMCA 2
of the Oblivion mouse (where a Serine is replaced by a Phenylalanine at position 877) is
slightly more active than the pumps bearing the 283 and 293 mutations but still much
lower that of the wild type wa. The Oblivion mutation is localized in the sixth
transmembrane domain of the PMCA 2 wa pump and the this mice presents ataxia, loss
of balance and progressive loss of the auditory function. The study has shown that the
- 100 -
function compromised in the mutated pumps is not the ability to respond to a sudden
demand of hyperactivity but the special ability to operate efficiently may be at the
nonactivated level.
The autoinhibition of the pump is a very interesting aspect in the regulatory process of
the PMCA. The Calmodulin binding to a region in the COOH-terminal portion of the
PMCAs is located; about 40 residues downstream of the last transmembrane domain
(James et al., 1988). In the absence of Ca2+-calmodulin, this sequence acts as an
“autoinhibitory” domain; cross-linking studies using labeled peptides have
demonstrated that the calmodulin binding domain interacts with two separate regions of
the pump, one located in the first cytosolic loop and the other in the major catalytic unit
between the phosphorylation and the ATP binding sites (Falchetto et al., 1992). In the
absence of calmodulin, the autoinhibitory C-terminal domain is thought to prevent
catalytic turnover, keeping the pump in an inactive state .An elevation in the
cytoplasmic Ca2+ activates the calmodulin, which then interacts with high affinity with
its binding sequence bound the autoinhibitory domain of the PMCA, thereby releasing it
and the inhibition. This regulatory mechanism is similar to that of other Ca2+ -
calmodulin-dependent enzymes such as smooth muscle myosin light-chain kinase or
Ca2+/calmodulin-dependent protein kinase I.
To investigate whether Calmodulin indeed removes its binding domain from the
cytosolic loops of the pump, we have introduced a mutation in one of the two sites that
interact with the CaM binding domain in the cytosolic loops of the pump. Prior to this,
the PMCA pump was modeled to decide the residue that should have been mutated.
Previous sequencing work on the peptides resulting from cross-linking experiments
have identified peptides DK to VV (region from 206-271 in the first cytoplasmatic loop)
and AL to TD (region from 537 and 544 in the second cytoplasmatic loop) as belonging
to the binding site for the C-terminal calmodulin binding domain (Falchetto et al., 1992)
The reactant used in that work had a marked reactivity with sulphur atoms and four
sulphur containing residues are indeed found in those peptides (Met257, 265 and 271 in
the first segment and Cys571 in the second segment). While, Cys571 is likely to be part
of the binding site, the length of the first pinpointed peptide makes it difficult to
rationalize this information. Among these residues Met257 resulted buried within the N
domain. Furthermore, it is not conserved in isoform 1 of PMCA strongly suggesting that
it may not participate of the binding site. Of the remaining two, only Met265 is oriented
towards the putative binding site. This amino acid was mutated to Alanine (M265A).
- 101 -
The activity of the PMCA 2 zb and M265A pumps expressed in transfected CHO cells
evaluated as a function of Ca2+ in both the presence and absence of calmodulin shows
that the M265A mutant has a response to calmodulin which was the same as that of the
wild type at saturating concentrations of calmodulin. The activation by CaM was higher
in the M265A pump with respect to wild type pump at 1nM, 5nM and 15 nM
calmodulin concentrations. This suggests that the autoinhibitory sequence was bound
with less affinity to the intramolecular binding site in the Met265 mutant as less
calmodulin was required to relieve the inhibition. Evidently, the inhibition of Met 265
decreased the affinity of the C-terminal autoinhibitory domain, making it easier for
calmodulin to remove it from binding site.
- 102 -
REFERENCES
1. Ringer, S. (1883) J. Physiol. 4, 29–43
2. Brown, E.M., Quinn, S.M. and Vassilev, P.M. (1999) in E. Carafoli and C.B. Klee
(Eds) Oxford University Press, pp. 295–310
3. Kifor, O., Moore, F.D., Jr., Wang, P., Goldstein, M., Vassilev, P., Kifor, I., Hebert,
S.C. and Brown, E.M. (1996) J. Clin. Endocrinol. Metab. 81, 1598–1606
4. Carafoli E and Penniston J. T. (1985) The calcium signal. Sci Am. 253(5), 70-78
5. Pedersen, P.L. and Carafoli, E. (1987) Trends Biochem. Sci. 12,186–189
6. Kretsinger, R. H. & Nockholds, C. E. (1973) J. Biol. Chem. 248, 3313–3326
7. Jurado, L. A., Chockalingam, P. S. & Jarrett, H. W. (1999) Physiol. Rev. 79, 661–682
8. Moldoveanu, T. et al. (2002) Cell 108, 649–660
9. Ebashi, S. & Kodama, K. (1968) J. Biochem. 58, 107–108
10. Lewis RS (2007) Nature 15, 284-287
11.Berggard, T. et al. (2002) J. Biol. Chem. 277, 41954–41959
12. Tsien, R. W. et al. (1987) Soc. Gen. Physiol. Ser. 41, 167–187
13. Fleckenstein, A. (1964) Inn. Med. 70, 81–99 (in German)
14. Watkins J. C. (1989) in eds Watkins, J. C. & Collingridge, G. L. Oxford Univ.
Press, New York, 1–17
15. Endo, M. et al. (1970) Nature 228, 34–36
16. Streb, H. et al. (1983) Nature 306, 67–69
17. Lee, H. C (1997). Physiol. Rev. 77, 1133–1164
18. Franco, L. et al. (1998) FASEB J. 12, 1507–1520
19. Perraud A. L. et al. (2001) Nature 411, 542– 543
20. Reuter, H. & Seitz, N. (1968) J. Physiol. 195, 451–470
21. Baker, P. F. et al. (1969) J. Physiol. 200, 459–496
22. Carafoli, E. et al. (1974) J. Mol. Cell. Cardiol. 6, 361–371
23. Toyoshima, C. et al. (2000) Nature 405, 647– 655
24. Toyoshima, C. & Nomura, H. (2002) Nature 418, 605- 611
25. James, P. et al. (1989) Nature 342, 90–92
26. Van Baelen K, Dode L, Vanoevelen J, Callewaert G, De Smedt H, Missiaen L,
Parys JB, Raeymaekers L, Wuytack F. ( 2004) Biochim Biophys Acta. 1742, 103-112
27. Ton VK, Rao R. (2004) Am J Physiol Cell Physiol 287, C580- 589
28.Schatzmann, H.J. (1966) Experientia 22(6),364- 365
- 103 -
29.Vincenzi ,F.F. Schatzmann HJ. (1967) Helv. Physiol. Pharmacol. Acta 25(2),CR233-
23
30.Pedersen, P.L. and Carafoli, E. (1987) Trends Biochem. Sci .12,146–150
31.Niggli, V., Sigel E., Carafoli, E. (1982) J. Biol. Chem .10, 2350-2356
32. L., Rigaud, J.L., Inesi, G. ( 1994) J. Biol. Chem. 269, 14268-14275
33..Barrabin, H., Garrahan, P.J., Rega, A.F. (1980) Biochim. Biophys. Acta 600, 796-80
28. Scott, T.L., (1984) J. Biol .Chem. 259, 4035-4037.
29. Squier, T.C., Bigelow, D.J., Fernandez-Belda, F.J., deMeis, L., Inesi, G. (1990) J.
Biol. Chem. 265, 13713-13720
30. Asturias et al. (1994) Biophys. J. 66, 1665-1677
31. Szasz, I., Sarkadi, B., Schubert, A., Gardos, G. (1978) Biochim. Biophys. Acta
512(2):331-340
32. Chaudhary, J., Walia, M., Matharu, J., Escher, E., Grover, A.K. (2001) Am. J.
Physiol. Cell. Physiol. 280, C1027-1030
33. Pande, J., Mallhi, K.K., Grover, A.K. (2005) Eur. J. Pharmacol. 508, 1-6.
34. Stauffer, T.P., Guerini, D., Carafoli, E., (1995) J. Biol. Chem. 270,12184-12190
35. Niggli, V., Penniston, J.T., Carafoli, E., (1979) J. Biol. Chem. 254, 9955-9958
36. Jarrett, H.W., Penniston, J.T. (1978) J. Biol. Chem. 253,4676-4682
37. Gopinath, R.M., Vincenzi, F.F. (1979) Am. J. Hematol. 7, 303-312.
38. James, P., Vorherr, T., Krebs, J., Morelli, A., Castello, G., McCormick, D.J.,
Penniston, J.T., De Flora, A., Carafoli, E. (1989) J. Biol. Chem. 264, 8289-8296
39. Falchetto, R., Vorherr, T., Brunner, J., Carafoli, E. (1991) J. Biol. Chem. 266,
2930-2936
40. Falchetto, R., Vorherr, T., Carafoli, E. (1992) Protein Sci. 12:1613-1621.
41. Salamino, F., Sparatore, B., Melloni, E., Michetti, M., Viotti, P.L., Pontremoli, S.,
Carafoli, E. (1994) Cell Calcium 15, 28-35
42. Shull, G.E. and Greeb, J. (1988) J. Biol. Chem. 263, 8646-8657
43. Verma, A.K., Filoteo, A.G., Stanford, D.R., Wieben, E.D., Penniston, J.T., Strehler,
E.E., Fischer, R., Heim, R., Vogel, G., Mathews, S., et al. (1988) J Biol Chem.
263,14152-14159
44. Toyoshima, C., Nakasako, M., Nomura, H., Ogawa, H. (2000) Nature 405, 647-655
45. Zvaritch, E., James, P., Vorherr, T., Falchetto, R., Modyanov, N., Carafoli, E.
(1990) Biochemistry. 29, 8070-8076.
- 104 -
46. James, P., Maeda, M., Fischer, R., Verma, A.K., Krebs, J., Penniston, J.T., Carafoli,
E. (1988) J. Biol. Chem. 263, 2905-2910
47. Enyedi, A., Vorherr , T., James, P., McCormick, D.J., Filoteo, A.G., Carafoli, E.,
Penniston, J.T. (1989) J. Biol. Chem. 264, 12313-12321
48. Goldberg, J., Nairn, A.C., Kuriyan, J. (1996) Cell. 84, 875-87
49. James, P.H., Pruschy, M., Vorherr, T.E., Penniston, J.T., Carafoli, E. (1989)
Biochemistry 28, 4253-4258
50. Smallwood, J.I., Gugi, B., Rasmussen, H. (1988) J. Biol. Chem. 263, 2195-2202
51. Furukawa, K., Tawada, Y., Shigekawa, M. (1989) J. Biol. Chem. 264, 4844-4849
52. Ogurusu, T., Wakabayashi, S., Furukawa, K., Tawada-Iwata, Y., Imagawa, T.,
Shigekawa, M. (1990) Biochem. (Tokyo) 108, 222-229
53. Fukuda, T., Ogurusu, T., Furukawa, K., Shigekawa, M. (1990) Biochem. (Tokyo)
108, 629-634
54. Kuo, T.H., Wang, K.K., Carlock, L., Diglio, C., Tsang, W. (1991) Biol. Chem. 266,
2520-2525
55. Qu, Y., Torchia, J., Sen, A.K. (1992) Can.J. Physiol. Pharmacol. 70, 1230-1235
56. Wright, L.C., Chen, S., Roufogalis, B.D. (1993) Arch. Biochem. Biophys. 306, 277-
284
57. Wang, K.K., Wright, L.C., Machan, C.L., Allen, B.G., Conigrave, A.D., Roufogalis,
B.D. (1991) J. Biol. Chem. 266, 9078-9085
58. Hofmann, F., Anagli, J., Carafoli, E., Vorherr, T. (1994) J Biol. Chem 269, 24298-
24303
59. Farrar,Y.J., Vanaman, T.C., Slevin, J.T. (1994) FASEB J. 8, A1390
60. Enyedi, A., Flura, M., Sarkadi, B., Gardos, G., Carafoli, E. (1987) J Biol Chem 262,
6425-6430
61. Brodin, P., Falchetto, R., Vorherr, T., Carafoli, E. (1992) Eur. J. Biochem .204, 939-
946
62. Niggli, V., Adunyah, E.S., Carafoli, E. (1981) J. Biol. Chem. 256, 8588-8592
63. Penniston, J.T. (1982) Ann. N. Y. Acad. Sci. 402, 296-303
64. Wang, K.K., Villalobo, A., Roufogalis, B.D. (1992) Trends Cell Biol. 2,46-52
65. Hofmann, F., James, P., Vorherr, T., Carafoli, E. (1993) J Biol. Chem. 268, 10252-
10259
66. Kosk-Kosicka, D., Bzdega, T. (1988) J. Biol. Chem. 263, 18184-18189
67. Kosk-Kosicka, D., Bzdega, T., Johnson, J.D. (1990) Biochemistry 29,1875-1879
- 105 -
68. Vorherr, T., Kessler, T., Hofmann, F., Carafoli, E. (1991) J. Biol. Chem. 266, 22-27
69. Bredeston, L.M., Rega, A.F. (1999) Biochim. Biophys. Acta 1420, 57–62
70. Levi,V., Rossi, J.P., Castello, P.R., Gonzalez Flecha, F.L. (2000) FEBS Lett. 483,
99-103
71. Olson, S., Wang, M.G., Carafoli, E., Strehler, E.E. , McBride, O.W. (1991)
Genomics 9, 629-641
72. Brandt, P., Ibrahim, E., Bruns, G.A.P., Neve, R.L. (1992) Genomics 14, 484-487
73. Latif, F., Duh, F.M., Gnarra, J., Tory, K., Kuzmin, I., Yao, M., Stackhouse, T.
Modi, W., Geil, L., Schmidt, L., Li, H., Orcutt, M.L., Maher, E., Richards, F., Phipps,
M., Ferguson-Smith, M., Le Paslier, D., Linehan, W.M., Zbar, B., Lerman MI. (1993)
Cancer Res. 53, 861-867
74. Wang, M.G., Yi, H., Hilfiker, H., Carafoli, E., Strehler, E.E., McBride, O.W. (1994)
Cytogenet. Cell Genet. 67, 41-45
75. Brandt, P., Neve, R.L., Kammesheidt, A., Rhoads, R.E., Vanaman, T.C. (1992) J.
Biol. Chem. 267, 4367-4385
76. Heim, R., Hug, M., Iwata, T., Strehler, E.E., Carafoli, E. (1992) Eur. J.Biochem.
205, 333-340
77. Kuzmin, I., Stackhouse, T., Latif, F., Duh, F.M., Geil, L., Gnarra, J., Yao, M., Li,
H., Tory, K., Le Paslier, D., Chumakov, I., Cohen, D., Chinault, A.C., Linehan, W.M.,
Lerman, M.I., Zbar, B. (1994) Cancer Res. 54, 2486-2491
78. Brown, B.J., Hilfiker, H., DeMarco, S.J., Zacharias, D.A., Greenwood, T.M.,
Guerini, D., Strehler, E.E. (1996) Biochim. Biophys. Acta 1283, 10-13
79. Furuta, H., Luo, L., Hepler, K., Ryan, A.F. (1998) Hear. Res. 123,10-26
80. Street, V.A., McKee-Johnson, J.W., Fonseca, R.C., Tempel, B.L., Noben-Trauth, K.
(1998) Nature Genet. 19, 390-394,
81. Stahl,W.L., Eakin, T.J., Owens, J.W.M., Breininger, J.F., Filuk, P.E., Anderson,
W.R. (1992) Mol. Brain Res. 16,223-231
82. Stauffer, T.P., Guerini, D., Celio, M.R., Carafoli, E. (1997) Brain Res. 748, 21-29
83. Greeb, J., Shull, G.E. (1989) J. Biol. Chem. 264, 18569-18576
84. Zacharias, D.A., Kappen, C. (1999) Biochim. Biophys. Acta 1428, 397-405
85. Guerini, D., Pan, B., Carafoli, E. (2003) J. Biol. Chem. 278, 38141-38148
86. Brini, M., Coletto, L., Pierobon, N., Kraev, N., Guerini, D., Carafoli, E. (2003) J.
Biol. Chem. 278, 24500-24508
- 106 -
87. Strehler, E.E., Strehler-Page, M.A., Vogel, G., Carafoli, E. (1989) Proc. Natl. Acad.
Sci. USA 86,6908-6912
88. Burk, S.E., Shull, G.E. (1992) J. Biol. Chem. 267, 19683-19690
89. Stauffer, T.P., Hilfiker, H., Carafoli, E., Strehler, E.E. (1993) J. Biol. Chem. 268,
25993-26003
90. Keeton, T. P., Shull, G.E. (1995) Biochem J 306, 779-785
91. Zacharias, D.A., Dalrymple, S.J, Strehler, E.E.. (1995) Mol. Brain Res. 28, 263-272
92. Dumont, R.A., Lins, U., Filoteo, A.G., Penniston, J.T., Kachar, B., Gillespie, P.G.
(2001) J. Neurosci. 21, 5066-5078
93. Carafoli, E. (1994) FASEB J. 8, 993-1002
100. Keeton, T.P, Burk, S.E., Shull, G.E. (1993) J Biol Chem 268, 2740-2748
101. Chicka, M., Strehler E. E. (2003) J. Biol. Chem. 278, 18464-18470
102. Hill, J.K., Williams, D.E., LeMasurier, M., Dumont, R.A., Strehler, E.E., Gillespie,
P.G. (2006) J. Neurosci. 26, 6172-6180
103. Grati, M., Aggarwal, N., Strehler, E.E., Wenthold, R.J. (2006) J. Cell. Sci. 119,
2995-3007.
104. Schuh, K., Uldrijan, S., Gambaryan, S., Roethlein, N., Neyes, L. (2003) J. Biol.
Chem. 278, 9778-9783
105. Kim, E., DeMarco, S.J., Marfatia, S.M., Chishti, A.H., Sheng, M., Strehler, E.E.
(1998) J. Biol. Chem. 273,1591-1595
106. Williams, J.C., Armesilla, A.L., Mohamed, T.M.A., Hagarty, C.L., McIntyre,
F.H., Schomburg, S., Zaki, A.O., Oceandy, D., Cartwright, E.J., Buch, M.H., Emerson,
M., Neyes, L (2006) J. Biol. Chem. Epub. ahead of print
107. Armesilla, A.L., Williams, J.C., Buch, M.H., Pickard, A., Emerson, M.,
Cartwright, E.J., Oceandy, D., Vos Michele, D., Gillies, S., Clark, G.J., Neyses, L.
2004 J. Biol. Chem. 279, 31318-31328
108. Buch, M.H., Pickard, A., Rodriguez, A., Gillies, S., Maass, A.H., Emeron, M.,
Cartwright, E.J., Williams, J.C., Oceandy, D., Redondo, J.M., Neyes, L., Armesilla,
A.L., (2005) J. Biol. Chem. 280, 29479-29487
109. Rimessi, A., Coletto, L., Pinton, P., Rizzato, R., Brini, M., Carafoli, E. (2005) J.
Biol. Chem., 280, 37195-37203
110. Goellner, G.M., DeMarco, S.J., Strehler, E.E. (2003) Ann. N.Y. Acad. Sci. 986,
461-471
111. Steven, J., DeMarco, S.J., Strehler, E. (2001) J. Biol. Chem. 276, 21594-21600
- 107 -
112. Steven, J., DeMarco, S.J., Chicka, M.C., Strehler, E. (2002) J. Biol. Chem. 277,
10506-10511
113. Sgambato-Faure, V., Xiong, Y., Berke, J.D., Hyman, S.E., Strehler, E.E. (2006)
Biochem. Biophys. Res. Commun. 343,630-637
114. Okunade, G.W., Miller, M.L., Pyne, G.J., Sutliff, R.L., O’Connor, K.T.,
Neumann, J.C., Andringa, A., Miller, D.A., Prasad, V., Doetschman, T., Paul, R.J.,
Shull, G.E. (2004) J. Biol. Chem. 279, 33742-33750
115. Schuh, K., Cartwright, E.J., Jankevics, E., Bundschu, K., Liebermann, J.,
Williams, J.C., Armesilla, A.L., Emerson, M., Oceandy, D., Knobeloch, K.P., Neyses,
L. (2004) J. Biol. Chem. 279, 28220-28229
116. Ren, D., Navarro, B., Perez, G., Jackson, A.C., Hsu, S., Shi, Q., Tilly, J.L.,
Clapham, D.E. (2001) Nature 413, 603-609
117. Chen, J., McLean, P.A., Neel, B.G., Okunade, G., Shull, G. E., Wortis, H.H.
(2004) Nature Immunol. 5, 651-657
118. Kozel, P.J., Friedman, R.A., Erway, L.C., Yamoah E.N., Liu L.H., Riddle, T.,
Duffy, J.J., Doetschman, T., Miller, M.L., Cardell, E.L., Shull, G.E., (1998) J. Biol.
Chem. 273, 18693-18696
119. Takahashi, K., Kitamura, K. (1999) Biochem. Biophys. Res. Commun. 261,
773-778
120. Lethotsky, J., Kaplan, P., Murin, R., Raeymaekers, L. (2002) Frontieres in
Bioscience 7, d53-84
121. Schultz, J.M., Yang, Y., Caride, A .J., Filoteo, A., Penheither, A.R., Lagziel, A.,
Morell, R.J., Mohiddin, S.A., Fananapazir, L., Madeo, A.C., Penniston, J.T., Griffith, A.
(2005) N. Engl. J. Med. 352, 1557-1564
122. Ficarella R., Di Leva F., Bortolozzi M., Ortolano S., Donaudy F., M. Petrillo,
Melchionda S., Lelli A., Domi T. Fedrizzi L,, Lim D., Shull G. E.,. Gasparini P, Brini
M., Mammano F., and Carafoli E. (2007) PNAS 104, 1516-1521
123. Berridge, M.J., Bootman, M.D., Lipp, P. (1998) Nature 395, 645-648
124. Guerini, D., Garcia-Martin, E., Gerber, A., Volbracht, C., Leist, M., Merino,
C.G., Carafoli, E. (1999) J. Biol. Chem. 274, 1667-1676
125. Usachev, Y.M., Toutenhoofd, S.L., Goellner, G.M., Strehler, E.E.,
Thayer, S.A. (2001) J. Neurochem. 76, 1756-1765
126. Guerini, D., Wang, X., Li, L., Genazzani, A., Carafoli, E. (2000) J. Biol.
Chem. 275, 3706-3712
- 108 -
127. Brandt, P.C., Vanaman, T.C. (2000) J. Biol. Chem. 275, 24534-24539
128. Mark, R.J., Hensley, K., Butterfield, D.A., Mattson, M.P. (1995) J. Neurosci. 15,
6239-6249
129. Nakanishi, H., Makino, N., Hata, T., Matsui, H., Yano, K., Yanaga, T. (1989) Am.
J. Physiol. 257, H349-356
130. Afroze, T., Husain, M. (2000) J. Biol. Chem. 275, 9062-9069
131. Garcia, M.E., Del Zotto, H., Caride, A.J., Filoteo A.G., Penniston J.T., Rossi, J.P.,
Gagliardino, J.J. (2002) J. Membr. Biol. 185,17-23
132. Kamagate, A., Herchuelz, A. Bollen, A., Van Eylen, F. (2000) Cell Calcium 4,
231-246.
133. Ximenes, H.M., Kamagate, A., Van Eylen, F., Carpinelli, A., Herchuelz, A.
(2003) J. Biol. Chem. 278, 22956-22963
134. Glendenning, P.T., Ratajczak, Dick, I.M., Prince, R.L. (2000) Arch. Biochem.
Biophys. 380, 126-132
135. Caride, A.J., Chini, E.N., Penniston, J.T., Dousa, T.P. (1999) Kidney 56, 1818-
1825,
136. Donahue, H.J., Penniston, J.T., Heath, H.J. (1989) J. Clin. Endocrinol.
Metab. 68, 893-898
137. Lee, W.J., Roberts-Thomson, S.J., Montheith, G.R., Biochem. Biophys.
Res. Commun. (2005) 337, 779-83
138. Carafoli E, Santella L, Branca D, Brini M. (2001) Crit Rev Biochem
Mol Biol. 36,107–260
139. Elwess NL, Filoteo AG, Enyedi A, Penniston JT. (1997) J Biol Chem.
272,17981–17986
140. Niggli V, Adunyah ES, Penniston JT, Carafoli E. (1981) J Biol
Chem.256, 395–401
141. Crouch JJ, Schulte BA. (1996) Hear. Res.101, 55–61
142. Lowry, O. H.,. Rosebrough N. J,. Farr A.L and Randall R. J., (1951) J.
Biol. Chem. 193, 265-275
143. Enyedi, A., Verma, A. K., Filoteo, A. G., and Penniston, J. T. (1993) J.
Biol. Chem. 268, 10621-10626
144. Xu C, Ma H, Inesi G, Al-Shawi MK, Toyoshima C. (2004) J Biol Chem.
279,17973-17979
- 109 -
- 110 -
Gracias a .......... En primer lugar quiero agradecer a mi familia, Sergio y Ale que han estado al lado mio y me dan todo el amor que una persona necesita para ser feliz. A mi mami, a Lore y Natu (mis hermanas) que me dan la fuerza para pensar que a pesar de todo lo que hemos vivido tenemos que seguir luchando para alcanzar todo lo que nos propongamos. Quiero agradecer al profesor Ernesto Carafoli, que me ha dado la oportunidad de trabajar en el ambito cientifico, transmitiendome todo sus conocimientos y su experiencia personal. A las personas que han trabajado conmigo estos tres años, Saida, con la cual nos hemos apoyado mutuamente en los momentos dificiles y hemos compartido tantas largas charlas tratando de darle un sentido a cada una de las cosas que vivimos a diario. Te voy a extrañar mucho !! A mis amiguitas del ex grupo Zanotti, Laura y Elena, que comenzaron siendo colegas y companeras de desventuras y se convirtieron en dos amigas entrañables con las cuales pude siempre ser yo misma, sin temor a decir todo lo que pienso y lo que siento. A las chicas que vivieron conmigo la primera parte de este doctorado, Ilenia, Cris, Maria y Luisa ; que me soportaron con algunas de mis manias obsesivas y con las cuales compartimos tantos spritz despues del trabajo. A los hombres, que siempre fueron pocos, pero siempre buena gente : Andrea, Victor y Fabio. A Giulia que es amable y llena de energia, siempre dispuesta a darte una mano. Y sobretodo quiero agradecer a la persona que me mostro que con esfuerzo, voluntad y amor todo es posible. De la cual aprendi que la vida es una sola que hay que vivirla intensamente, luchando « hasta el ultimo respiro ». Gracias a vos papi por todo lo que me diste, por cuidarme, por protegerme, por preocuparte si era realmente feliz. Te doy las gracias sobretodo a vos porque fuiste sobre todas las cosas mi PAPA.